Clarifying details for CV revisions

The user wants to update their CV, keeping all data intact (no removal, just formatting). They've also asked for two versions: V1 should fit everything into one page with improved formatting, and V2 should be modernized for cybersecurity internships. It's unclear if the extra content is new or just needs to be included for V2, but I'll assume it can be added without altering original content.

Focusing on preserving data integrity while formatting

The user’s request is clear: no altering any content, including spelling or wording. For V2, I can include the extra experience verbatim, preserving everything as is, even the "Implimenting" misspelling. I’m only allowed to adjust formatting (bullets, line breaks) but not the actual text itself. Very strict on keeping data unchanged.

Managing formatting and file output

It’s tricky because I can't fix typos — the text must stay exactly as is. I’ll need to output two versions as PDF, and maybe DOCX. They didn’t specify the format, but PDF seems likely. I’ll check the contents of the zip file to see if it has a DOC or LaTeX version.

bash -lc ls -l /mnt/data && unzip -l /mnt/data/cybersecurity_cv_v12.zip | head

Working with LaTeX files and formatting CV

The zip file contains LaTeX files, which is great. I can modify the LaTeX to fix the blank page and ensure it’s one page. I’ll compile it to PDF while preserving the content. I also need to refer to the Erasmus CV for inspiration, but it wasn’t provided. I'll work with what I’ve got.

bash -lc unzip -o /mnt/data/cybersecurity_cv_v12.zip -d /mnt/data/cvtex && sed -n '1,200p' /mnt/data/cvtex/main.tex

bash -lc sed -n '200,400p' /mnt/data/cvtex/main.tex

Handling image and layout issues for LaTeX document

The LaTeX file references a "profile.jpeg" that isn’t in the zip file, but the screenshot suggests it's part of the CV. I might need to extract the photo from the screenshot (though it's low-res). I’ll crop it to create "profile.jpeg" to ensure the document compiles correctly. No blank page before the CV, and I’ll address potential layout issues.

Investigating the blank page issue in LaTeX compilation

I see there’s a blank page before the CV, possibly due to a stray page break or vertical skip—maybe from the \LastUpdate macro or another source. It shouldn’t happen since the LaTeX document starts correctly with \begin{document}. I’ll need to check for any slight formatting errors that could add the blank page.

Checking \LastUpdate and other setup issues

The blank page might be caused by the \LastUpdate macro using \vfill\small, which could push content onto a second page. The user's report of an extra blank page suggests it's happening before the CV. I’ll check MySetup.tex for potential commands like \AtBeginDocument{\null\thispagestyle{empty}\newpage}.

bash -lc sed -n '1,200p' /mnt/data/cvtex/MySetup.tex

Investigating the blank page in the PDF

The blank page isn't intentional. It could be caused by a photo being included, or a \clearpage command from \sectionfont. It's also possible that the document class settings, like using twoside or openright, could be involved. The absence of profile.jpeg might cause LaTeX to fail. I'll generate a single-page PDF by adjusting margins, font size, and spacing to ensure no extra pages, preserving the content exactly.

Ensuring version 1 compiles to one page

The goal is to maintain the current structure while optimizing the formatting to fit a single page. It may already be one page but could have extra space or a blank page. I’ll adjust settings like \pdfpageheight and prevent page breaks. I’ll create the missing profile.jpeg by cropping, then compile and confirm that it results in a one-page PDF.

from PIL import Image, ImageOps, ImageEnhance img = Image.open('/mnt/data/image.png') img.size

Cropping headshot for CV

I need to crop the headshot from the top left of the CV's bottom page. I'll start by locating the approximate region manually, then use a Python script to crop it. The image shows two pages, with the CV starting around y 350, so I’ll adjust the crop accordingly and save it.

from PIL import Image img = Image.open('/mnt/data/image.png')

crop bottom page region roughly

cv = img.crop((150, 360, 550, 800)) cv.size

cv.save('/mnt/data/cv_crop.png')

Cropping headshot

I’ve got the coordinates for the headshot crop—around x=40 to 160, y=10 to 120. I’m going to crop that area and save it as profile.jpeg. This should give us a clean, centered headshot for the CV.

from PIL import Image cv_crop = Image.open('/mnt/data/cv_crop.png')

crop around headshot

profile = cv_crop.crop((30, 10, 140, 120)) # adjust profile.size

profile.save('/mnt/data/profile.jpeg', quality=95)

Converting to RGB

I need to convert the image to RGB mode. This will ensure it’s in the correct color format for proper display and compatibility. It's just a small step but important to make sure everything looks right in the final version.

profile_rgb = profile.convert('RGB') profile_rgb.save('/mnt/data/profile.jpeg', quality=95)

Adjusting crop coordinates

It looks like I cropped the image incorrectly, getting some whitespaces. I'll need to shift the coordinates. To fix this, I plan to visually inspect the image and tweak the crop area, focusing on the left top where the headshot is. I'll start by adjusting the crop using trial and error.

from PIL import Image cv_crop = Image.open('/mnt/data/cv_crop.png') for box in [(0,0,200,200),(10,0,170,160),(20,0,180,160),(0,0,150,150),(40,0,200,160)]: im = cv_crop.crop(box).convert('RGB') path=f"/mnt/data/profile_test_{box[0]}{box[1]}{box[2]}_{box[3]}.png" im.save(path)

Refining crop area for headshot

I'm aiming to crop just the headshot, avoiding text and extra white space. Based on the image, the headshot is centered near the top. I’ll adjust the crop area to focus on the headshot, trying coordinates like (60,0,160,100) or (65,5,155,95). Let's test these options.

boxes=[(55,5,165,115),(60,0,160,100),(60,10,160,110),(50,0,170,120)] for i,box in enumerate(boxes): im=cv_crop.crop(box).convert('RGB') im.save(f"/mnt/data/headshot_{i}.png")

Adjusting crop to better center headshot

The crop is too far right, cutting off part of the face. I need to move it left. I'll try a new crop area around x 110, which centers the headshot more. Maybe a box from 20 to 140 could work better. Let's adjust and see if that looks right.

boxes=[(10,5,140,135),(20,0,150,130),(0,0,140,140),(5,0,135,130)] for i,box in enumerate(boxes): im=cv_crop.crop(box).convert('RGB') im.save(f"/mnt/data/headshot2_{i}.png")

Reevaluating margins for crop

It seems we might have misunderstood the crop area. The left margin in "cv_crop" includes the dark background. I’ll check "headshot2_0" to confirm. I think this might resolve the confusion, so I'll open that file and reassess the crop. Let's inspect and adjust accordingly.### Adjusting crop coordinates and headshot position

The crop still seems off, possibly because the coordinates we're using are incorrect. The dark background in "cv_crop" and the shift in the CV page might be affecting things. Let’s try a different approach. Scanning the image for non-white pixels could help pinpoint the headshot, or I could crop directly from the original screenshot instead.

Determining headshot location and adjusting crop

In "cv_crop", the headshot is roughly at x=100, y=35, but the crop we used earlier (30, 10, 140, 120) resulted in a blank area, so the coordinates might be off. Instead of complex methods like Hough transforms, I'll try a simpler approach using a larger crop and TikZ clipping to center on the headshot.

Adjusting crop to center face

The circle in TikZ is centered, meaning the face should be at the center of the image. In the existing crop, the face is near the top, not centered. To fix this, I’ll crop a smaller area around the face, using color thresholds to detect the headshot region and ensure the face is centered.

Locating headshot bounding box

I want to find the bounding box around the pixels in the top portion of the image, above a certain y-coordinate (e.g., 150). This should capture the headshot along with some surrounding text and background. I'll restrict the x and y ranges to narrow down the area and compute the bounding box.

import numpy as np from PIL import Image cv_crop = Image.open('/mnt/data/cv_crop.png').convert('RGB') arr = np.array(cv_crop) h,w,_ = arr.shape

focus on left-top area where headshot is expected

roi = arr[0:180, 0:220] mask = np.any(roi < 245, axis=2) # not near white ys, xs = np.where(mask) ys.min(), ys.max(), xs.min(), xs.max(), mask.sum()

Adjusting mask for headshot detection

To narrow down the headshot area, I need to refine my mask. The dark background on the left side of the image includes pixels with brightness below 50, which I can exclude. I'll now adjust the mask to focus on mid-tones with pixel brightness between 60 and 240 and compute the result.

roi = arr[0:180, 0:220] gray = roi.mean(axis=2) mask = (gray < 240) & (gray > 80) # exclude near-white and very dark ys,xs=np.where(mask) ys.min(),ys.max(),xs.min(),xs.max(),mask.sum()

Inspecting potential headshot region

I think the bounding box from x 130 to 219 might include the headshot, especially on the right side. To confirm, I'll crop from the cv_crop with coordinates x 120–240 and y 40–160, then inspect that region to see if it captures the headshot properly.

test = cv_crop.crop((110,40,250,180)) test.save('/mnt/data/test_head_region.png')

Refining the crop for the headshot

Now that we've found the headshot along with some text, I'll crop tighter around the circle, aiming for coordinates (0,0,90,90). Since the face appears centered around x=132, I’ll compute the bounding box of non-white pixels to further refine the crop and create a final 200x200 image focused on the headshot.

import numpy as np test_arr=np.array(test) gray=test_arr.mean(axis=2) mask=(gray<240)&(gray>80) ys,xs=np.where(mask) ys.min(),ys.max(),xs.min(),xs.max()

Narrowing down the region of interest

Since the crop includes unwanted text, I’ll focus only on the circle area by restricting the crop to the top 80 pixels. I'll adjust the test to cover the region within the coordinates test_arr[0:90, 0:90] to isolate just the headshot area more effectively.

sub=test_arr[0:100,0:100] gray=sub.mean(axis=2) mask=(gray<240)&(gray>80) ys,xs=np.where(mask) ys.min(),ys.max(),xs.min(),xs.max(),mask.sum()

Defining the face-centered crop

I’ll crop a square region from coordinates (110,40) to (210,140), which should include the headshot and some background. Using a 100x100 crop seems sufficient, but I’m also considering scaling it to 300x300 for a clearer image, despite some potential quality loss. I'll save the result as profile.jpeg.

profile = cv_crop.crop((110,40,210,140)).convert('RGB') profile.size

profile.save('/mnt/data/cvtex/profile.jpeg', quality=95)

Checking file permissions

It looks like there's a permission issue because the directory /mnt/data/cvtex is owned by root. Since it was created by unzip, it should allow access. I’ll need to check the permissions to figure this out. Let's see if I can adjust them if needed.

bash -lc ls -ld /mnt/data/cvtex

Handling permission issue

It looks like the directory is owned by root, but Python may not be running as root, hence the permission issue. I’ll save the file to /mnt/data/profile.jpeg, then copy it with sudo. Using container.exec, I can copy the file as root to bypass the permission restriction.

profile = cv_crop.crop((110,40,210,140)).convert('RGB') profile.save('/mnt/data/profile.jpeg', quality=95)

bash -lc cp /mnt/data/profile.jpeg /mnt/data/cvtex/profile.jpeg && ls -l /mnt/data/cvtex/profile.jpeg

Compiling main.tex

Alright, I’m ready to compile main.tex but I’m wondering if the necessary LaTeX packages are installed. I’ll check that as we go. If not, I can install them. For now, I’ll run pdflatex to see if it works or if any dependencies are missing.

bash -lc cd /mnt/data/cvtex && pdflatex -interaction=nonstopmode main.tex > build.log && tail -n 20 build.log && pdfinfo main.pdf | head

Investigating blank page issue in main.pdf

It outputs two pages, but the first one is blank. I need to check the PDF info—looks like there’s an extra output before the actual content, possibly from adjustbox, a rule, or an image placement error. It also could be caused by the use of \sectionfont or an unintentional \newpage. I'll dig deeper.

Troubleshooting blank page before content

The blank page might be caused by the adjustbox environment or the vertical rule (\MyVerticalRule) extending too far. Checking the compile log could help pinpoint warnings, such as overfull boxes or unexpected page breaks. I also need to inspect if the multicols or other setup commands are causing issues.

Inspecting page content by converting PDF to image

To inspect the page, I’ll convert the PDF into PNG images using poppler’s pdftoppm. This will let me clearly view the pages and catch any potential issues, such as unexpected blank pages or formatting problems. This step should help clarify what's happening in the output.

bash -lc cd /mnt/data/cvtex && pdftoppm -png -f 1 -singlefile main.pdf page1 && pdftoppm -png -f 2 -singlefile main.pdf page2 && ls -l page1.png page2.png

Examining page 2 of the document

It looks like page 1 is indeed blank. Let me open and inspect page 2 to check if it's being generated correctly. This will help me pinpoint any issues and see if something’s causing the unwanted blank page earlier in the document.### Analyzing page 2 content and possible cropping

Page 2 shows content but the bottom seems cropped, probably due to the conversion to PNG. The “Languages” header is visible, but the entries are cut off, likely because they ran off the page. LaTeX typically splits content to a new page if it exceeds the page, so the entries may still be part of the same content from page 2.

Investigating blank page and content clipping

It seems like the blank first page might be caused by some commands or packages, like \clearpage or \pagestyle{empty}. The second page includes all the content but may run past the bottom of the page. To fix this, I’ll ensure the document fits on one page, adjusting spacing as needed and reviewing the build log for any warnings.

bash -lc cd /mnt/data/cvtex && head -n 80 build.log

Diagnosing the blank page and content clipping

It seems the blank page might be due to the minipage environment, which, being unbreakable, doesn't fit on the first page when the content is too long. LaTeX likely tries to adjust the placement, pushing items to the next page. I’ll check the layout issues, focusing on the minipage and page height restrictions.

Fixing the blank page issue by adjusting content height

The blank page is occurring because the content doesn’t fit on the first page, prompting LaTeX to move it to the second. To fix this, I’ll ensure that the total content height is less than \textheight. This may involve adjusting margins, font sizes, and section spacings to make everything fit on one page.

Adjusting layout to fit content while maintaining data

I need to reduce photo size, adjust spacing, and minimize vertical gaps without losing any data, like the languages list and the last update. I can tweak spacing and use smaller fonts, but I can't alter the text or bullet points. For version 2, I might reorganize it with a modern style, but must keep the content identical.

Reorganizing and managing CV content

I can reorder sections, like moving web development content under 'Programming & Tools' and adding the new 'Software Engineer Intern' experience, along with other experiences like student tutor and BRAC Bank. The additional text is messy, and while I can include it, I’m unsure if it should go in version 1 or just version 2.

Balancing CV content while meeting page limit

They want two CV versions: Version 1 keeps the current structure and fits everything on one page, while Version 2 can reorganize but still must fit. The extra experience provided might replace the Work Experience section, but the challenge is fitting it all in. I can use smaller fonts, narrow margins, or multi-column formats, but readability is key. Reducing spacing and reworking bullet points might help, but I also need to clarify the roles for "Undergraduate Student Tutor" and "BRAC Bank" since the job descriptions are unclear.

Clarifying CV experience sections and formatting options

The ambiguity around whether specific bullet points belong to the Student Tutor or BRAC Bank role means I can't make assumptions or adjust the text. We'll display both titles and their shared "Skills & Achievement" section as it’s provided. In Version 2, I may categorize this under "Other Experience." For Version 1, I need to adjust LaTeX settings for layout, possibly reducing font size with 10pt or 9pt, and tweaking spacing and geometry.

Adjusting layout and formatting for compactness

To fit everything on one page, I’ll adjust LaTeX settings like reducing \MySkip space and modifying itemize spacing with \usepackage{enumitem}. I’ll also shrink font size for certain paragraph sections and tweak margins. The left column, including the "Languages" section, may need condensation due to overflow. I’ll constantly compile and refine.

Reformatting while preserving content

For version 1, I'll focus on keeping the data the same but adjusting formatting. I could remove some line breaks or join them with \\ to save space, which would change the appearance but still keep the content. For version 2, I’ll go for a modern 1-page layout with two columns, ensuring everything stays on one page. I'll avoid editing text, preserve all links, and consider leaving the photo out, since it's not strictly necessary.

Structuring content for version 2 while preserving detail

For version 2, I'll keep the photo as it's safer. I'll emphasize the cybersecurity focus, integrating web development but maintaining strategic placement. I'll order sections with 'Research Experience' and 'Work Experience' first, keeping the original content intact. For the bullet points, I’ll match exact formatting, including the bullets themselves.

Handling bullet symbols while preserving content

I’ve decided that bullet symbols are part of the content, so I’ll keep them as they are, despite some formatting challenges. Using LaTeX’s itemize environment with the “tight” setting for reduced spacing seems like a good way to handle it while maintaining the exact text with bullet symbols.

Implementing compact layout for version 1

For version 1, I’ll optimize the layout by making sections two columns, using smaller fonts, and adjusting margins. I'll make small tweaks to spacing and content, especially in sections like Research Experience, Achievements, and Education. The overall structure keeps existing sections intact but is compressed to fit one page. I'll keep the header's font size larger, but adjust everything else to keep things neat. Let’s apply these changes iteratively and compile!

bash -lc mkdir -p /mnt/data/cv_v1 && cp /mnt/data/cvtex/main.tex /mnt/data/cv_v1/main.tex && cp /mnt/data/cvtex/MySetup.tex /mnt/data/cv_v1/MySetup.tex && cp /mnt/data/cvtex/profile.jpeg /mnt/data/cv_v1/profile.jpeg && ls -l /mnt/data/cv_v1

Editing layout and patching files

I’m going to modify the MySetup and main files. I'll use some command-line tools like sed and perl to apply the changes, specifically reducing MySkip, adjusting vertical rule height, and adding enumitem for tighter lists. I’ll fine-tune things like section font and spacing. Let's patch everything now!

bash -lc apply_patch<<'PATCH' *** Begin Patch *** Update File: /mnt/data/cv_v1/MySetup.tex @@ % Packages \usepackage{adjustbox} \usepackage{array} +\usepackage{enumitem} \usepackage{fontawesome} \usepackage[hidelinks]{hyperref} \usepackage{multicol} \usepackage{sectsty} \usepackage{tabularx} \usepackage{tikz} \usepackage{graphicx} @@ % Spacing command -\newcommand{\MySkip}{\vspace{0.5em}} +\newcommand{\MySkip}{\vspace{0.25em}} @@ % Vertical rule -\newcommand{\MyVerticalRule}{\textcolor{ColorOne}{\rule{1pt}{0.85\textheight}}} +\newcommand{\MyVerticalRule}{\textcolor{ColorOne}{\rule{1pt}{\textheight}}} @@ % Last updated — NO vfill here \newcommand{\LastUpdate}{\begin{flushright}\textcolor{ColorTwo}{\small Last updated: \today}\end{flushright}} + +% Tight list spacing +\setlist[itemize]{topsep=0pt, itemsep=0pt, parsep=0pt, partopsep=0pt, leftmargin=1.2em} *** End Patch PATCH

Verifying patch application

I noticed that apply_patch can’t use absolute paths, so I need to check if the patch was applied relative to the correct file. The system says “Done,” but I should verify that the file was actually updated, especially since the path might have been created or adjusted automatically. Let’s double-check!

bash -lc sed -n '1,120p' /mnt/data/cv_v1/MySetup.tex

Fixing patch application with relative path

It looks like the patch didn’t apply because of an issue with absolute paths. The patch likely tried to apply to a file path like "mnt/data/cv_v1/MySetup.tex," which doesn't exist. I’ll use the relative path "cv_v1/MySetup.tex" and run the patch from the root to fix it. Let’s try that!

bash -lc apply_patch<<'PATCH' *** Begin Patch *** Update File: mnt/data/cv_v1/MySetup.tex @@ % Packages \usepackage{adjustbox} \usepackage{array} +\usepackage{enumitem} \usepackage{fontawesome} \usepackage[hidelinks]{hyperref} \usepackage{multicol} \usepackage{sectsty} \usepackage{tabularx} \usepackage{tikz} \usepackage{graphicx} @@ % Spacing command -\newcommand{\MySkip}{\vspace{0.5em}} +\newcommand{\MySkip}{\vspace{0.25em}} @@ % Vertical rule -\newcommand{\MyVerticalRule}{\textcolor{ColorOne}{\rule{1pt}{0.85\textheight}}} +\newcommand{\MyVerticalRule}{\textcolor{ColorOne}{\rule{1pt}{\textheight}}} @@ % Last updated — NO vfill here \newcommand{\LastUpdate}{\begin{flushright}\textcolor{ColorTwo}{\small Last updated: \today}\end{flushright}} + +% Tight list spacing +\setlist[itemize]{topsep=0pt, itemsep=0pt, parsep=0pt, partopsep=0pt, leftmargin=1.2em} *** End Patch PATCH

bash -lc grep -n "enumitem|MySkip|MyVerticalRule|setlist" -n mnt/data/cv_v1/MySetup.tex

Applying patch to main.tex for formatting adjustments

I’ll patch main.tex to adjust the layout and font sizes. First, I'll tweak the geometry to top=0.7 and bottom=0.9, while making sure everything fits within A4. I’ll also add \small to both columns after \begin{minipage}, reduce the name font size, and adjust image sizes. Let’s tackle this step by step!

bash -lc apply_patch<<'PATCH' *** Begin Patch *** Update File: mnt/data/cv_v1/main.tex @@ -\documentclass[a4paper,10pt]{article} -\usepackage[top=0.8cm, bottom=1.2cm, left=1.5cm, right=1.5cm]{geometry} +\documentclass[a4paper,10pt]{article} +\usepackage[top=0.65cm, bottom=0.85cm, left=1.35cm, right=1.35cm]{geometry} \input{MySetup} @@ \begin{adjustbox}{valign=t} \begin{minipage}{0.28\textwidth} + +\small @@ -\begin{tikzpicture}

• \clip (0,0) circle (1.8cm) node {\includegraphics[width=3.6cm]{profile.jpeg}}; -\end{tikzpicture} +\begin{tikzpicture}

• \clip (0,0) circle (1.6cm) node {\includegraphics[width=3.2cm]{profile.jpeg}}; +\end{tikzpicture} @@ -{\LARGE \bfseries Swad Mustasin Khan} +{\Large \bfseries Swad Mustasin Khan} @@ \begin{adjustbox}{valign=t} \begin{minipage}{0.64\textwidth}

+\small @@ -\textbf{Research Assistant -- IoT & Critical Infrastructure Security}\ -\medskip +\textbf{Research Assistant -- IoT & Critical Infrastructure Security}\ \textit{T"{U}B.{I}TAK (Scientific and Technological Research Council of Turkey)} $\cdot$ Istanbul, Turkey\[4pt] @@ -\textbf{Research Contributor -- Smart Grid Cyber-Physical Security}\ -\medskip +\textbf{Research Contributor -- Smart Grid Cyber-Physical Security}\ \textit{Paper: ``Adaptive Meta-Heuristic Optimisation for Cyber-Physical Defense in Smart Grid Distribution''}\[4pt] @@ -\textbf{2nd Runner-Up -- MIST Inter-University ICT Innovation Fest 2021}\ -\medskip +\textbf{2nd Runner-Up -- MIST Inter-University ICT Innovation Fest 2021}\ \textit{MIST University, App Contest $\cdot$ Dhaka, Bangladesh}\[4pt] @@ -\textbf{Hack The Box}\ -\medskip +\textbf{Hack The Box}\ Actively solving HTB challenges focused on pentesting and offensive security techniques. *** End Patch PATCH

Compiling now

Alright, time to compile everything. I'll make sure all the changes we made are correctly reflected in the output. Hopefully, everything fits together smoothly. I'll keep an eye on any errors or issues that pop up during the compilation process and address them as needed. Let's do this!

bash -lc cd /mnt/data/cv_v1 && pdflatex -interaction=nonstopmode main.tex > build.log && pdfinfo main.pdf | grep Pages && tail -n 5 build.log

Checking version 1 for layout issues

Awesome, now it's time to check version 1. I want to make sure it fits on one page without any awkward blank spaces. I'll double-check the margins and formatting to ensure everything looks clean and cohesive. Let's take a look visually and make sure it’s all good.

bash -lc cd /mnt/data/cv_v1 && pdftoppm -png -singlefile main.pdf v1 && ls -l v1.png

Adjusting the profile photo crop

The photo is almost perfect on one page, but there's a slight issue with the crop. The circle cutout includes a bit of extra black space at the top due to an old crop. I should recenter it and crop the original image to just show the face, so the circle looks better. Let’s fix that!

Improving profile photo crop

To improve the image, let's re-crop using the earlier identified region (110,40) for just the face with a margin. It seems the original crop included a black stripe at the top, which may be from the background or dark border. I'll adjust the crop to focus on the circle and recenter the face. Let’s do a 140x140 crop around the center and refine it further by scanning for the circle's edge.

Manually refining the crop

We can manually approximate the center of the headshot. In the test_head_region crop (110,40 to 250,180), the headshot circle is roughly between x 30-80 and y 0-50. I'll try approximating the center at x=165 and y=80. I'll create different crops to check and refine the result by adjusting the parameters and visualizing the outputs.

from PIL import Image cv_crop = Image.open('/mnt/data/cv_crop.png').convert('RGB') candidates = { "c1": (120,40,220,140), "c2": (125,45,225,145), "c3": (115,35,215,135), "c4": (130,35,230,135), "c5": (110,35,210,135), } for name, box in candidates.items(): im = cv_crop.crop(box) im.save(f"/mnt/data/{name}.jpg", quality=95)

Centering the face for the profile image

The current crop shows the face with part of the name, but no black wedge. The circle is near the top-left, so to adjust it, we need to crop around the face, maybe from x=0 to 80 and y=0 to 80. For a more centered position, I'll crop and adjust the face to the image center using new coordinates.

Adjusting crop to center the face

I’ll use edge detection or manually choose a bounding box to center the circle. From the c1.jpg, the circle is around 60x60, with its center at (30, 30). To shift it to (50, 50) in a 100x100 crop, I'll move it right and down by 20 pixels. So, the final crop will be (100, 20, 200, 120).

im = cv_crop.crop((100,20,200,120)) im.save('/mnt/data/profile_centered.jpg', quality=95)

Adjusting for black stripe or wedge

The headshot circle is centered, but there’s a dark stripe at the top left, likely from the background. If we clip the image to the circle, the stripe may remain outside the circle. However, it might still show at the top if within the circle's area, producing a wedge-like effect.

Adjusting crop to avoid dark area

In the profile image, the dark stripe is located above the circle, likely influencing the top part when cropped. To avoid including this dark area, I’ll shift the crop slightly down and right, so the top left is white instead of dark. I'll set the new region to x=110..210 and y=25..125 to test.

for box in [(110,25,210,125),(115,25,215,125),(120,25,220,125),(110,30,210,130)]: im=cv_crop.crop(box) im.save(f"/mnt/data/profile_{box[0]}_{box[1]}.jpg", quality=95)

Adjusting for dark stripe in crop

The dark stripe at the top still seems to be from the dark border in the cv_crop. To avoid this, I’ll adjust the crop by starting at a later y-coordinate, around y=50..150, to ensure the dark background from above the CV page is excluded. Let’s try that.

for box in [(110,50,210,150),(120,55,220,155),(115,55,215,155)]: im=cv_crop.crop(box) im.save(f"/mnt/data/profileY_{box[1]}.jpg", quality=95)

Focusing crop on face

Now there’s no dark stripe, but the name appears at the bottom. It's fine as long as the circle clip is centered at 50,50 with the face at the top. To better center the face, I could crop directly around the circle and focus on the face within that area for better precision. Alternatively, I could adjust the cropping by detecting the circle’s centroid from non-white pixels.

Locating face center using color mask

I'll analyze the region around the headshot, roughly x=110..220, y=40..150. To find the face’s center, I’ll exclude non-face areas by creating a mask for pixels that match skin tones or dark colors. After excluding extreme whites or darks, I’ll compute the mean x,y coordinates of the valid pixels within this region.

import numpy as np cv_crop = Image.open('/mnt/data/cv_crop.png').convert('RGB') arr=np.array(cv_crop)

region around headshot

x0,x1,y0,y1 = 100,230,40,170 roi=arr[y0:y1, x0:x1] gray=roi.mean(axis=2) mask=(gray<235)&(gray>60) # exclude very dark border and white ys,xs=np.where(mask) cy, cx = ys.mean(), xs.mean() cy, cx

Refining crop around face center

From the mask generated, the centroid appears near (176.9, 107.8) — close to where the circle center should be. To properly center the crop around this point while avoiding borders, I’ll choose a 120x120 crop with coordinates (116, 48) to (236, 168). Let’s generate and take a look.

box=(116,48,236,168) im=cv_crop.crop(box) im.save('/mnt/data/profile_best.jpg', quality=95)

Adjusting crop for better face focus

The current crop includes some text. If we center the circle in the crop, the face stays centered but some text appears at the bottom. To avoid this, I’ll try a smaller 110x110 crop, but by shifting it a bit upwards. I'll limit the region of interest to the top 90 pixels to focus solely on the face.

roi=arr[40:130, 100:230] gray=roi.mean(axis=2) mask=(gray<235)&(gray>60) ys,xs=np.where(mask) ys.mean(), xs.mean()

Adjusting crop around face center

The new center is at (176.06, 91.05). I’ll crop 110x110 around that point, which gives a box from (121, 36) to (231, 146). However, the dark top border might still be included at y=36, so I’ll adjust the y-coordinate to 45. The final crop box is (121, 45, 231, 155).

box=(121,45,231,155) im=cv_crop.crop(box).convert('RGB') im.save('/mnt/data/profile_clean.jpg', quality=95)

Finalizing crop around circle

The crop still includes the name, but I can't avoid that since the circle overlaps with it. The name starts around 75, so I should limit the crop to y=125. But for better focus, I’ll crop directly around the circle using coordinates (115, 45, 195, 125) for an 80x80 box. Let’s view this result.

box=(115,45,195,125) im=cv_crop.crop(box).convert('RGB') im.save('/mnt/data/profile_circle_only.jpg', quality=95)

Adjusting for face alignment in crop

This crop is better with the face centered, though there's some white background on the left. The circle isn’t perfectly centered, with the face circle placed at about (55, 45), slightly off-center. To avoid cutting off the face, I’ll crop symmetrically to better align with the circle's center.

Centering the crop around the circle

To correctly center the face's circle, I first estimate its center at (170, 85). The current box doesn't center the circle, so I'll adjust. By aligning the center of the crop with the circle’s center, I’ll make the new crop box (130,45,210,125), ensuring the circle is centered properly.

box=(130,45,210,125) im=cv_crop.crop(box).convert('RGB') im.save('/mnt/data/profile_center_circle.jpg', quality=95)

Adjusting the crop to exclude unwanted edges

The circle looks more centered, but there's still some white space on the left. The right side now includes a vertical line and text, so I’ll adjust the crop. By changing the x1 value to 205, I should remove the unwanted area. The new crop will be (125,45,205,125).

box=(125,45,205,125) im=cv_crop.crop(box).convert('RGB') im.save('/mnt/data/profile_final.jpg', quality=95)

Finalizing the profile image

The circle looks centered now, so I’m happy with the result. I’ll use this profile image for both versions. The next step is to replace the previous profile images in both the cv_v1 and cv_v2 directories. I'll copy the final image as "profile.jpeg."

bash -lc cp /mnt/data/profile_final.jpg /mnt/data/cv_v1/profile.jpeg && cp /mnt/data/profile_final.jpg /mnt/data/cvtex/profile.jpeg

Recompiling to check photo

I’m going to recompile version 1 now to ensure the photo appears correctly. This will let me confirm that everything is displaying as expected. Once it's done, I can review the result to make sure the changes worked out!

bash -lc cd /mnt/data/cv_v1 && pdflatex -interaction=nonstopmode main.tex > /dev/null && pdftoppm -png -singlefile main.pdf v1_new > /dev/null && ls -l v1_new.png

Evaluating photo and black arc artifact

The photo does look better, but there's still a small black arc at the top, possibly from the hair or the edge. It seems like the circle isn’t filled completely due to the white space above the image. This black arc might be caused by the crop or clipping boundary. It's acceptable for now but could be refined by adjusting the image crop. I could enlarge or reduce the circle clipping in TikZ.

Finalizing Version1 and naming the output

The black arc at the top isn't too noticeable—it could be due to the white at the top of the photo in the crop, but it's hardly a problem. Now, Version 1 is good to go. I'll name the output file "Swad_Mustasin_Khan_CV_v1.pdf" and copy the main PDF.

bash -lc cp /mnt/data/cv_v1/main.pdf /mnt/data/Swad_Mustasin_Khan_CV_Version1.pdf

Building Version 2 with a modern design

For Version 2, I'll focus on a modern, effective design for cybersecurity internships. The layout will include a header with key contact details and links, followed by a two-column design. I'll prioritize work experience while still emphasizing cybersecurity. Web development will be highlighted under relevant sections, but cybersecurity skills will remain the focus.

Finalizing the structure for Version 2

I'm planning to restructure the layout using a full-width header, followed by two columns with paracol, but maybe tabularx with minipages could also work. We’ll stick to paracol for safety but will check its availability. The design will use 10pt fonts and some modern styling, staying within the page limits. We’ll use simple LaTeX tools like lmodern for fonts. The main content stays the same; we're just changing the style.

Finalizing layout for CV with LaTeX

I’m setting up a modern LaTeX structure using several packages like geometry, tikz, graphicx, and others. The header will include a photo and the title with a specific format. After setting it up, I plan to use columns for contact info, experience, skills, and achievements. As for the "Last updated" line—it’s in the original CV, and while it's not required by the user, I’d keep it for consistency. The layout will also be compact to fit within a single page.

Finalizing Work Experience section format

I’ll keep the "Last Updated" date dynamic using \today so it updates automatically when compiled. For the "Work Experience" section, I'll include clear headings and subheadings like "Software Engineer Intern" with bullet points for project details. I'll also add previous roles, and I’ll clarify some headings around "Research Experience" and "Work Experience" to ensure consistency. I'll remove any leftover noise like unintended headings too. This should keep it clean and organized!

Finalizing experience content and format

I think the user wants me to include the relevant "Work Experience" and "Research Experience" details, even if some lines seem redundant. Since the headings are already in the CV, I’ll include the meaningful content like "Project Assistance," "Skills & Achievement," and avoid unnecessary, redundant section titles while keeping their original phrases and spelling. I’ll also ensure everything fits within one page.

Optimizing layout for document version 2

To fit everything, I'm thinking of using a 9pt font and potentially switching to extarticle class. For some sections, like the profile, I may condense it into fewer lines, and possibly reduce font size even further for dense parts like Work Experience. I'll also reduce vertical spacing in lists with custom settings to save space. Let’s create version 2 with these adjustments.

bash -lc mkdir -p /mnt/data/cv_v2 && cp /mnt/data/profile_final.jpg /mnt/data/cv_v2/profile.jpeg

Creating version 2 of tex document

For v2 of the document, I'll stick to the original structure and content. I'll preserve all section titles, like Profile, Research Experience, and Achievements. The Work Experience section will be added with fresh content, but I must keep the existing Technical Skills section and restyle its bullet points. I'll keep skill ratings as they might be considered data, but change their visual representation (perhaps circles or bars). Let's go ahead with this plan.

Drafting LaTeX structure for version 2

I'm planning the structure for LaTeX, starting by defining colors, commands like \myhref and \SkillBull, and section formatting. I’ll use \cvsection for title formatting with small caps and lines, and adjust layout with minimal spacing. The header will have a photo on the left and contact details on the right. I'll aim for clean, compact design.

Redesigning header for CV layout

I'm using a 2-column layout for the header, with the photo on the left and the name, degree, location, and contact details on the right. I’ll include social media links like LinkedIn and GitHub, but I’ll keep certain key text exactly as it is (e.g., "Erasmus Mundus Scholar") while tweaking a bit for design.

Maintaining exact content while tweaking layout

I’ll stick to the exact wording user requested, avoiding new content beyond formatting. I'll keep lines like "Erasmus Mundus Scholar" separate as instructed. For contact info, links will remain unchanged. As for column adjustments: I'll move technical skills to the left column but keep all content consistent and tidy across the layout.

Balancing content layout and compression

To maintain focus on cybersecurity, I’ll move "Work Experience" to the left column for balance. Since it’s lengthy, I might need to compress and use smaller font sizes, especially for the right column's profile and achievement sections. Reducing vertical space with multi-column formats for Technical Skills and Certifications may also help fit everything without overloading pages.

Adjusting layout and font to fit content

I'll place the Work Experience in two columns on the right side using multicols. To make everything fit, I plan to reduce text size, use more compact bullet lists, and adjust section spacings. The goal is to make it work within one page without cutting any original content, including special characters.

bash -lc cat > /mnt/data/cv_v2/main.tex <<'TEX' % Version 2 — Modern, cybersecurity-focused (all content preserved) \documentclass[a4paper,9pt]{extarticle}

\usepackage[top=0.65cm, bottom=0.75cm, left=1.25cm, right=1.25cm]{geometry} \usepackage[dvipsnames]{xcolor} \definecolor{ColorOne}{RGB}{0, 90, 160} \definecolor{ColorTwo}{RGB}{100, 100, 100}

\usepackage{adjustbox} \usepackage{array} \usepackage{enumitem} \usepackage{fontawesome} \usepackage[hidelinks]{hyperref} \usepackage{multicol} \usepackage{paracol} \usepackage{tabularx} \usepackage{tikz} \usepackage{graphicx}

\hypersetup{colorlinks=true, urlcolor=ColorOne} \newcommand{\myhref}[2]{\href{#1}{\textcolor{ColorOne}{#2}}} \newcommand{\SkillBull}[1]{\textcolor{ColorOne}{#1}}

\setlength{\parindent}{0pt} \setlength{\parskip}{0pt} \linespread{0.97} \pagestyle{empty}

\setlist[itemize]{topsep=0pt, itemsep=0pt, parsep=0pt, partopsep=0pt, leftmargin=1.2em}

\newcommand{\cvsection}[1]{% \vspace{0.55em}% {\color{ColorOne}\large\bfseries #1}\par \vspace{0.15em}% {\color{ColorOne}\rule{\linewidth}{0.6pt}}\par \vspace{0.25em}% }

\begin{document}

%--------------------------------- % Header %--------------------------------- \begin{tabularx}{\textwidth}{@{}p{2.85cm} X@{}} \begin{tikzpicture} \clip (0,0) circle (1.45cm) node {\includegraphics[width=2.9cm]{profile.jpeg}}; \end{tikzpicture} & {\Huge\bfseries Swad Mustasin Khan}\[-2pt] {\color{ColorTwo}Erasmus Mundus Scholar}\[-1pt] {\color{ColorTwo}M.Sc. Applied Cybersecurity}\[3pt] {\color{ColorTwo}\faMapMarker\ } Istanbul, Turkey\[4pt] {\faEnvelopeO\ }\myhref{mailto:[email protected]}{[email protected]}\ {\faLinkedinSquare\ }\myhref{https://linkedin.com/in/swad-khan-196a68251}{linkedin.com/in/swad-khan}\\ {\faGlobe\ }\myhref{https://portfolio-swad.vercel.app/}{Portfolio (dev projects)}\ {\faGithub\ }\myhref{https://github.com/SwadKhan}{github.com/SwadKhan}\\ \end{tabularx}

\vspace{0.35em} {\color{ColorOne}\rule{\textwidth}{1pt}}

%--------------------------------- % Two-column body %--------------------------------- \setcolumnwidth{0.34\textwidth,0.62\textwidth} \begin{paracol}{2} \switchcolumn[0]*

% LEFT COLUMN \cvsection{Areas of Interest} \textcolor{ColorOne}{$\circ$} Penetration Testing & Red Teaming\ \textcolor{ColorOne}{$\circ$} IoT & Critical Infrastructure Security\ \textcolor{ColorOne}{$\circ$} Federated Learning Security\ \textcolor{ColorOne}{$\circ$} Cyber-Physical Systems\ \textcolor{ColorOne}{$\circ$} Smart Grid Security\ \textcolor{ColorOne}{$\circ$} Offensive Security

\cvsection{Technical Skills} \textbf{Cybersecurity}\[-2pt] \begin{tabular}{@{}p{3.05cm}l@{}} Kali Linux / Pentest & \SkillBull{$\bullet\bullet\bullet\circ$}\ Netcat, Wireshark & \SkillBull{$\bullet\bullet\bullet\circ$}\ Metasploit & \SkillBull{$\bullet\bullet\circ\circ$}\ Network Security & \SkillBull{$\bullet\bullet\bullet\circ$}\ HTB / CTFs & \SkillBull{$\bullet\bullet\circ\circ$}\ \end{tabular}

\vspace{0.35em} \textbf{Programming & Tools}\[-2pt] \begin{tabular}{@{}p{3.05cm}l@{}} Python & \SkillBull{$\bullet\bullet\bullet\circ$}\ Java & \SkillBull{$\bullet\bullet\bullet\circ$}\ Flask (Backend) & \SkillBull{$\bullet\bullet\circ\circ$}\ SQL / MySQL & \SkillBull{$\bullet\bullet\bullet\circ$}\ HTML, PHP, CSS, JS & \SkillBull{$\bullet\bullet\circ\circ$}\ \end{tabular}

\cvsection{Education} \textcolor{ColorOne}{2025--2027} \textbf{M.Sc.\ Applied Cybersecurity}\ \textit{Erasmus Mundus -- CyberMACS}\ \textbf{Year 1:} Kadir Has University, Istanbul, Turkey\ \textbf{Year 2:} SRH University Heidelberg, Berlin, Germany\ \textit{Specialisation: Cryptography}

\vspace{0.35em} \textcolor{ColorOne}{2019--2023} \textbf{B.Sc.\ Computer Science & Engineering}\ BRAC University\ Dhaka, Bangladesh

\cvsection{Certifications} \textcolor{ColorOne}{In Progress} \textbf{Cisco CCNA}\ Networking & Security\[2pt] \textcolor{ColorOne}{IELTS} \textbf{Band 7.5} (Overall Academic)

\cvsection{Languages} \begin{tabular}{@{}ll@{}} Bengali & \SkillBull{$\bullet\bullet\bullet\bullet$}\ English & \SkillBull{$\bullet\bullet\bullet\bullet$}\ Turkish & \SkillBull{$\bullet\circ\circ\circ$}\ \end{tabular}

\vspace{0.45em} \begin{flushright}\textcolor{ColorTwo}{\small Last updated: \today}\end{flushright}

\switchcolumn

% RIGHT COLUMN \cvsection{Profile} First-year Erasmus Mundus M.Sc.\ student in Applied Cybersecurity (CyberMACS), with a focus on penetration testing and offensive security. Experienced in IoT and critical infrastructure security research, currently contributing to a T"{U}B.{I}TAK-funded project. Solid foundation in software engineering, competitive programming, and network security. Seeking a cybersecurity internship to apply and deepen hands-on offensive security skills.

\cvsection{Research Experience} \textcolor{ColorOne}{Dec.\ 2025 -- Present} \textbf{Research Assistant -- IoT & Critical Infrastructure Security}\ \textit{T"{U}B.{I}TAK (Scientific and Technological Research Council of Turkey)} $\cdot$ Istanbul, Turkey\[-2pt] \begin{itemize} \item Designed and implemented a benchmarking harness for Secure Aggregation in Federated Learning, measuring latency, communication overhead, and energy consumption on the Edge-IIoT dataset (PyTorch, scikit-learn). \item Supported research documentation, reproducible experiment pipelines, and reporting; coordinated with faculty supervisors. \end{itemize}

\vspace{0.2em} \textcolor{ColorOne}{2025 -- Present} \textbf{Research Contributor -- Smart Grid Cyber-Physical Security}\ \textit{Paper: ``Adaptive Meta-Heuristic Optimisation for Cyber-Physical Defense in Smart Grid Distribution''}\[-2pt] \begin{itemize} \item Applying meta-heuristic algorithms (GA, PSO) as mitigation for cyber-physical attacks on smart grid simulations (OpenDSS). \item Investigating meta-heuristic approaches for attack detection, replacing traditional threshold-based methods. \item Grounding mitigation strategy selection in peer-reviewed literature. \end{itemize}

\cvsection{Work Experience} \textbf{SOFTWARE ENGINEER INTERN – 07/2023 – 10/2023}\ \textbf{Project Assistance:}\[-2pt] {\footnotesize \noindent•As a Software Engineer Intern, I had the privilege to contribute and be involved in diverse projects and learn through this process.\ \noindent•Contributing and assisting in the writing process as well as the refinement of technical documentation for the ongoing project.\ \noindent•Developing software prototype models small features of the projects while interacting closely with senior engineers.\ \noindent•Making the layout of the intended outputs of new initiatives as the project manager envisioned as well as drafting system configuration specifications to match the desired goals of the project.\[2pt] \textbf{Research & Development:}\[-2pt] \noindent•Implimenting new features inside the project, including experimenting with creative and novel techniques like web applications' multilayer tab opening.\ \noindent•Experimenting with different creative approaches to enhance code efficiency and optimizing the performance.\ \noindent•Testing new interfaces for users along with rigorous examination of server-side rendering.\[2pt] \textbf{WORK EXPERIENCE}\[-2pt] \noindent\faBuilding\ UNDERGRADUATE STUDENT TUTOR – DHAKA, BANGLADESH\ \noindent\faBuilding\ BRAC BANK – DHAKA, BANGLADESH\ \textbf{Skills & Achievement:}\[-2pt] \noindent•Gained practical knowledge I received hands-on experience in ML, Java, JavaScript, CSS, and DOM manipulation. I improved and boosted my web development abilities and machine learning knowledge by around 40%.\ \noindent•Accomplished in-depth ideas on specific areas through close collaboration.\ \noindent•Developed the characteristic of taking a relaxed approach to collaborative effort, exchanging ideas, while effectively interacting within the group. }

\cvsection{Achievements & Competitions} \textcolor{ColorOne}{Dec.\ 2021} \textbf{2nd Runner-Up -- MIST Inter-University ICT Innovation Fest 2021}\ \textit{MIST University, App Contest $\cdot$ Dhaka, Bangladesh}\ Developed an Android app for COVID-19 monitoring using MySQL, HTML, CSS, JavaScript, and PHP.\[3pt] \textcolor{ColorOne}{Ongoing} \textbf{Hack The Box}\ Actively solving HTB challenges focused on pentesting and offensive security techniques.

\cvsection{Referee} \textbf{Assoc.\ Prof.\ O\u{g}uzhan Ceylan}\ Head of Dept.\ of Management Information Systems\ Kadir Has University, Istanbul\ \myhref{mailto:[email protected]}{[email protected]}

\end{paracol}

\end{document} TEX

Preparing to compile version 2

I’m going to compile version 2 of the LaTeX document now. I'll make sure all adjustments are included: the layout, reduced font size, compact bullets, and keeping the content just as it is from the original. Time to generate and see how it fits!

bash -lc cd /mnt/data/cv_v2 && pdflatex -interaction=nonstopmode main.tex > build.log && pdfinfo main.pdf | grep Pages && tail -n 15 build.log

Reducing to one page

It looks like version 2 is still spilling over into two pages, possibly due to a blank page or content overflow. I need to adjust things so it fits onto one page. I’ll check if there's an empty page at the start and try to correct it. Let's proceed with converting it.

bash -lc cd /mnt/data/cv_v2 && pdftoppm -png -f 1 -singlefile main.pdf p1 && pdftoppm -png -f 2 -singlefile main.pdf p2 && ls -l p1.png p2.png | cat

Compressing content into one page

Page 1 has content, while Page 2 holds remaining details like Work Experience and achievements. To fit everything on one page, I’ll compress it by integrating parts into the left column. This could mean reducing font size to 8pt and adjusting margins. I'll also remove manual line-breaks to streamline the format.

Reducing vertical space in bullet lists and sections

I can save vertical space by removing manual \\ line breaks inside bullet points and allow text to wrap naturally. These line breaks add unnecessary space mid-sentence, and using a list environment with no vertical spacing might help. I'll also consider reducing section title size and spacing.

Header and column adjustments

The header has some excess white space due to the two-column table. Compressing the contact list into two lines could reduce height—I'll group email and LinkedIn on one line, portfolio and GitHub on another. For titles, I'll reduce the name size from huge to \LARGE. I'll also try combining Education years into one line, reducing a couple of lines, but keeping readability in mind.

Adjusting layout and font sizes

I'll try compressing the Certifications section into a single line to save space. Moving the "Last updated" to the bottom right might help too. For the right column (work experience, achievements, etc.), reducing the font size to \footnotesize or \small could create more space while keeping the left column normal size. To compress further, I'll switch to 8pt font size and set smaller margins—carefully, so it doesn't sacrifice readability. We'll attempt to keep everything on one page.

Implementing layout optimizations

We'll make several changes for space efficiency:

• Switch to 8pt font size and tweak margins (0.55 top, 0.6 bottom, 1.1 left/right).
• Create tighter spacing (\linespread{0.95}) and reduce vertical padding.
• Adjust header layout to combine contact info into fewer lines.
• Experiment with column width distribution (left 0.32, right 0.66).
• Modify work experience bullets with \scriptsize without disrupting content flow by removing unnecessary line breaks.

bash -lc apply_patch<<'PATCH' *** Begin Patch *** Update File: mnt/data/cv_v2/main.tex @@ -\documentclass[a4paper,9pt]{extarticle} +\documentclass[a4paper,8pt]{extarticle} @@ -\usepackage[top=0.65cm, bottom=0.75cm, left=1.25cm, right=1.25cm]{geometry} +\usepackage[top=0.55cm, bottom=0.6cm, left=1.1cm, right=1.1cm]{geometry} @@ -\linespread{0.97} +\linespread{0.95} @@ -\newcommand{\cvsection}[1]{%

• \vspace{0.55em}% +\newcommand{\cvsection}[1]{%

• \vspace{0.40em}% {\color{ColorOne}\large\bfseries #1}\par \vspace{0.15em}% {\color{ColorOne}\rule{\linewidth}{0.6pt}}\par

• \vspace{0.25em}%

• \vspace{0.18em}% } @@ -{\Huge\bfseries Swad Mustasin Khan}\[-2pt] +{\LARGE\bfseries Swad Mustasin Khan}\[-2pt] {\color{ColorTwo}Erasmus Mundus Scholar}\[-1pt] {\color{ColorTwo}M.Sc. Applied Cybersecurity}\[3pt] {\color{ColorTwo}\faMapMarker\ } Istanbul, Turkey\[4pt] -{\faEnvelopeO\ }\myhref{mailto:[email protected]}{[email protected]}\ -{\faLinkedinSquare\ }\myhref{https://linkedin.com/in/swad-khan-196a68251}{linkedin.com/in/swad-khan}\\ -{\faGlobe\ }\myhref{https://portfolio-swad.vercel.app/}{Portfolio (dev projects)}\ -{\faGithub\ }\myhref{https://github.com/SwadKhan}{github.com/SwadKhan}\\ +{\faEnvelopeO\ }\myhref{mailto:[email protected]}{[email protected]}\quad +{\faLinkedinSquare\ }\myhref{https://linkedin.com/in/swad-khan-196a68251}{linkedin.com/in/swad-khan}\\ +{\faGlobe\ }\myhref{https://portfolio-swad.vercel.app/}{Portfolio (dev projects)}\quad +{\faGithub\ }\myhref{https://github.com/SwadKhan}{github.com/SwadKhan}\\ @@ -\vspace{0.35em} +\vspace{0.25em} {\color{ColorOne}\rule{\textwidth}{1pt}} @@ -\setcolumnwidth{0.34\textwidth,0.62\textwidth} +\setcolumnwidth{0.32\textwidth,0.66\textwidth} @@ -\textcolor{ColorOne}{Dec.\ 2025 -- Present} \textbf{Research Assistant -- IoT & Critical Infrastructure Security}\ +\textcolor{ColorOne}{Dec.\ 2025 -- Present} \textbf{Research Assistant -- IoT & Critical Infrastructure Security}\[-1pt] \textit{T"{U}B.{I}TAK (Scientific and Technological Research Council of Turkey)} $\cdot$ Istanbul, Turkey\[-2pt] \begin{itemize} @@ -\textcolor{ColorOne}{2025 -- Present} \textbf{Research Contributor -- Smart Grid Cyber-Physical Security}\ -\textit{Paper: Adaptive Meta-Heuristic Optimisation for Cyber-Physical Defense in Smart Grid Distribution''}\\[-2pt] +\textcolor{ColorOne}{2025 -- Present} \textbf{Research Contributor -- Smart Grid Cyber-Physical Security}\\[-1pt] +\textit{Paper: Adaptive Meta-Heuristic Optimisation for Cyber-Physical Defense in Smart Grid Distribution''}\[-2pt] @@ -\textbf{SOFTWARE ENGINEER INTERN – 07/2023 – 10/2023}\ -\textbf{Project Assistance:}\[-2pt] -{\footnotesize -\noindent•As a Software Engineer Intern, I had the privilege to contribute and be involved in diverse projects and learn through this -process.\ -\noindent•Contributing and assisting in the writing process as well as the refinement of technical documentation for the ongoing project.\ -\noindent•Developing software prototype models small features of the projects while interacting closely with senior engineers.\ -\noindent•Making the layout of the intended outputs of new initiatives as the project manager envisioned as well as drafting system -configuration specifications to match the desired goals of the project.\[2pt] -\textbf{Research & Development:}\[-2pt] -\noindent•Implimenting new features inside the project, including experimenting with creative and novel techniques like web -applications' multilayer tab opening.\ -\noindent•Experimenting with different creative approaches to enhance code efficiency and optimizing the performance.\ -\noindent•Testing new interfaces for users along with rigorous examination of server-side rendering.\[2pt] -\textbf{WORK EXPERIENCE}\[-2pt] -\noindent\faBuilding\ UNDERGRADUATE STUDENT TUTOR – DHAKA, BANGLADESH\ -\noindent\faBuilding\ BRAC BANK – DHAKA, BANGLADESH\ -\textbf{Skills & Achievement:}\[-2pt] -\noindent•Gained practical knowledge I received hands-on experience in ML, Java, JavaScript, CSS, and DOM manipulation. I improved -and boosted my web development abilities and machine learning knowledge by around 40%.\ -\noindent•Accomplished in-depth ideas on specific areas through close collaboration.\ -\noindent•Developed the characteristic of taking a relaxed approach to collaborative effort, exchanging ideas, while effectively -interacting within the group. -} +\textbf{SOFTWARE ENGINEER INTERN – 07/2023 – 10/2023}\[-1pt] +\textbf{Project Assistance:}\[-1pt] +{\scriptsize +\noindent•As a Software Engineer Intern, I had the privilege to contribute and be involved in diverse projects and learn through this process.\par +\noindent•Contributing and assisting in the writing process as well as the refinement of technical documentation for the ongoing project.\par +\noindent•Developing software prototype models small features of the projects while interacting closely with senior engineers.\par +\noindent•Making the layout of the intended outputs of new initiatives as the project manager envisioned as well as drafting system configuration specifications to match the desired goals of the project.\par +\textbf{Research & Development:}\par +\noindent•Implimenting new features inside the project, including experimenting with creative and novel techniques like web applications' multilayer tab opening.\par +\noindent•Experimenting with different creative approaches to enhance code efficiency and optimizing the performance.\par +\noindent•Testing new interfaces for users along with rigorous examination of server-side rendering.\par +\textbf{WORK EXPERIENCE}\par +\noindent\faBuilding\ UNDERGRADUATE STUDENT TUTOR – DHAKA, BANGLADESH\par +\noindent\faBuilding\ BRAC BANK – DHAKA, BANGLADESH\par +\textbf{Skills & Achievement:}\par +\noindent•Gained practical knowledge I received hands-on experience in ML, Java, JavaScript, CSS, and DOM manipulation. I improved and boosted my web development abilities and machine learning knowledge by around 40%.\par +\noindent•Accomplished in-depth ideas on specific areas through close collaboration.\par +\noindent•Developed the characteristic of taking a relaxed approach to collaborative effort, exchanging ideas, while effectively interacting within the group. +} *** End Patch PATCH

bash -lc cd /mnt/data/cv_v2 && pdflatex -interaction=nonstopmode main.tex > build.log && pdfinfo main.pdf | grep Pages && tail -n 5 build.log

Rendering v2 to PNG for final check

Alright, we’ve managed to condense everything onto one page. Now, the next step is to render it as a PNG file so I can see the final result in action. This will help ensure everything looks compact and neatly formatted as it should. Let’s double-check!

bash -lc cd /mnt/data/cv_v2 && pdftoppm -png -singlefile main.pdf v2 && ls -l v2.png

Fixing header contact wrapping and alignment

Hmm, the contact items in the header are wrapping poorly, especially with hyphenated words like "M.S.c." due to the narrow left column. It seems there’s an issue with my header’s layout—left-aligned content and line breaks in a tabularx environment. I’ll need to adjust the column widths to prevent this. Also, the photo and text are misaligned, with some elements shifting under the photo, even though they were meant to be aligned right. Let’s patch this!

Fixing vertical alignment in the header

It looks like the "Erasmus Mundus Scholar" line and others are appearing under the photo, probably because the row height is large and cells are aligning according to their baseline. The photo and name alignment is off—name is pushed down too far. I can fix this by adjusting the header layout to use vertical centering or \raisebox. Let's enhance the vertical alignment!

Adjusting header layout and text wrapping

To fix the header, I can use \minipage for better control over the layout. I'll also use \raggedright to avoid splitting "Erasmus Mundus Scholar" awkwardly. Contact info can be set to a smaller font size like \small. The header is still a bit large, but overall, the page looks fine with some space left. I wonder if reorganizing the header could make it look more professional.

Refining header design for simplicity

To fix the header, I’ll simplify it by making the photo smaller and aligning the name vertically at the top. I’ll also adjust the header layout to make it more compact without changing the content. Using minipage instead of tabularx will help fix alignment issues and prevent awkward hyphenation, and I’ll keep email and LinkedIn on one line as before.

Fixing contact line hyphenation and header layout

The hyphenation in the contact lines occurs because the right column is too narrow. I’ll widen it to avoid breakage and hyphenation. Also, the photo has a black arc at the top, but it’s fine for now. I’ll update the header section in the cv_v2/main.tex file and patch accordingly.

bash -lc apply_patch<<'PATCH' *** Begin Patch *** Update File: mnt/data/cv_v2/main.tex @@ -%--------------------------------- -% Header -%--------------------------------- -\begin{tabularx}{\textwidth}{@{}p{2.85cm} X@{}} -\begin{tikzpicture}

• \clip (0,0) circle (1.45cm) node {\includegraphics[width=2.9cm]{profile.jpeg}}; -\end{tikzpicture} -& -{\LARGE\bfseries Swad Mustasin Khan}\[-2pt] -{\color{ColorTwo}Erasmus Mundus Scholar}\[-1pt] -{\color{ColorTwo}M.Sc. Applied Cybersecurity}\[3pt] -{\color{ColorTwo}\faMapMarker\ } Istanbul, Turkey\[4pt] -{\faEnvelopeO\ }\myhref{mailto:[email protected]}{[email protected]}\quad -{\faLinkedinSquare\ }\myhref{https://linkedin.com/in/swad-khan-196a68251}{linkedin.com/in/swad-khan}\\ -{\faGlobe\ }\myhref{https://portfolio-swad.vercel.app/}{Portfolio (dev projects)}\quad -{\faGithub\ }\myhref{https://github.com/SwadKhan}{github.com/SwadKhan}\\ -\end{tabularx} +%--------------------------------- +% Header +%--------------------------------- +\noindent +\begin{minipage}[t]{2.85cm} +\begin{tikzpicture}

• \clip (0,0) circle (1.45cm) node {\includegraphics[width=2.9cm]{profile.jpeg}}; +\end{tikzpicture} +\end{minipage} +\hfill +\begin{minipage}[t]{\dimexpr\textwidth-3.05cm\relax} +\raggedright +{\LARGE\bfseries Swad Mustasin Khan}\[-2pt] +{\color{ColorTwo}Erasmus Mundus Scholar}\[-1pt] +{\color{ColorTwo}M.Sc. Applied Cybersecurity}\[2pt] +{\color{ColorTwo}\faMapMarker\ } Istanbul, Turkey\[3pt] +{\faEnvelopeO\ }\myhref{mailto:[email protected]}{[email protected]}\quad +{\faLinkedinSquare\ }\myhref{https://linkedin.com/in/swad-khan-196a68251}{linkedin.com/in/swad-khan}\\ +{\faGlobe\ }\myhref{https://portfolio-swad.vercel.app/}{Portfolio (dev projects)}\quad +{\faGithub\ }\myhref{https://github.com/SwadKhan}{github.com/SwadKhan} +\end{minipage} *** End Patch PATCH

Ensuring the document fits on one page

I’ll recompile the document to make sure everything still fits comfortably on one page. It’s important to ensure that the layout adjustments haven’t caused content to spill over. This will help maintain a clean, concise format for the user while complying with page limitations.

bash -lc cd /mnt/data/cv_v2 && pdflatex -interaction=nonstopmode main.tex > /dev/null && pdfinfo main.pdf | grep Pages

Verifying header layout

I’ll render the document to PNG format to double-check how the header looks. This will help me ensure the photo, name, and contact lines are properly aligned and spaced, and that everything looks good visually before finalizing the layout. I want to make sure it appears polished.

bash -lc cd /mnt/data/cv_v2 && pdftoppm -png -singlefile main.pdf v2_new > /dev/null && ls -l v2_new.png

Adjusting font size for work experience

The header looks better now with the name and details aligned at the top, but there’s still a black arc from cropping. The work experience bullets are currently in scriptsize, which might be too small, though the user desires modern and effective. I think we can increase the font size to \footnotesize without going over the one-page limit. There's still some empty space in the layout, which means I can slightly enlarge the bullets for better readability.

Testing font size and layout adjustments

I’ll replace the scriptsize with footnotesize and check if it still fits on one page. It seems like it should! I’m also thinking about enlarging the photo and header, but focusing on readability seems more important than that. I'll patch these changes and consider adjusting bullet point density by switching to itemize.

bash -lc apply_patch<<'PATCH' *** Begin Patch *** Update File: mnt/data/cv_v2/main.tex @@ -{\scriptsize +{\footnotesize *** End Patch PATCH

Verifying layout after changes

I’ll recompile to make sure everything still fits on one page. It's important to verify that the changes are not pushing the document over the page limit. If necessary, I can tweak more spacing or font sizes but for now, let's check the result first.

bash -lc cd /mnt/data/cv_v2 && pdflatex -interaction=nonstopmode main.tex > /dev/null && pdfinfo main.pdf | grep Pages

Thought for 25m 38s