MASTER CV TEMPLATE - Mohammad Bilal
How to use: This is your single source of truth. For each application, copy this file, keep the CONSTANT blocks, and pick ONE variant from each SWAP block matching the role track (tracks:
AI= AI/LLM/GenAI roles,ML= ML/Data Science roles,FS= Full-Stack/Frontend/Backend/Web,CV= Computer Vision,GEN= Graduate SWE / everything else). Delete all instruction lines (anything starting with>) before exporting to PDF. ATS rules: single column, no tables/icons/graphics, standard fonts, exact keywords from the job description, every acronym written both ways once - e.g. "RAG (Retrieval-Augmented Generation)".
HEADER - CONSTANT
MOHAMMAD BILAL AI/ML & Software Engineer bilalnadeema302003@gmail.com | +974 55470358 | Doha, Qatar LinkedIn: linkedin.com/in/mohammad-bilal-64489827b | GitHub: github.com/Bixal99 | Portfolio: portfolio-bixal99s-projects.vercel.app
If the job is remote-international, add: "Open to remote (UTC+3) and relocation." If the job is Qatar onsite, add: "Based in Doha - available immediately, no relocation needed."
PROFESSIONAL SUMMARY - SWAP (pick one, then inject 2–3 exact phrases from the job description)
[AI] - AI Engineer / LLM Application Developer / Generative AI Developer Computer Science graduate building production AI applications with LLMs (Large Language Models), RAG (Retrieval-Augmented Generation), and agentic workflows. Hands-on with LangChain, LangGraph, OpenAI, Gemini, and Claude APIs across deployed projects including a voice-enabled AI book assistant and an AI-native SaaS for engineering teams. Strong Python and TypeScript foundations for shipping end-to-end AI products.
[ML] - Machine Learning Engineer / Data Scientist Computer Science graduate with end-to-end machine learning experience: data preprocessing, feature engineering, model training (XGBoost, scikit-learn), evaluation, and deployment via interactive Streamlit dashboards. Built a complete telecom churn prediction system with explainable AI-driven retention insights. Skilled in Python, Pandas, NumPy, and data visualization (Plotly).
[FS] - Full-Stack / Frontend / Backend / Web Developer Computer Science graduate and full-stack developer with six deployed production web applications built with React.js, Next.js, TypeScript, and Node.js. Experienced with REST APIs, PostgreSQL, MongoDB, and modern UI development (Tailwind CSS, Framer Motion). Collaborating engineer on an AI-native SaaS product, comfortable owning features from Figma design to Vercel deployment.
[CV] - Computer Vision Engineer Computer Science graduate specialized in real-time computer vision with OpenCV and MediaPipe. Built a hands-free eye-blink-to-Morse-code communication system using EAR-based blink detection and facial landmark tracking. Strong Python, NumPy, and image processing fundamentals, with additional experience in OCR (Optical Character Recognition) and diffusion models.
[GEN] - Graduate Software Engineer / generic Computer Science graduate (FAST NUCES, 2026) with strong fundamentals in data structures, algorithms, and OOP (Object-Oriented Programming), and hands-on experience across AI, machine learning, and full-stack development. Ships real products: six deployed web apps, an ML pipeline, and computer vision systems. Daily problem-solving practice (LeetCode) and collaborative Git workflow.
EXPERIENCE - CONSTANT (this is your strongest credential - never omit)
Software Engineer (Collaborating Engineer) - Zennore / Halyo · Remote · Oct 2025 – Present
- Contributing engineer on Halyo, an AI-native SaaS workspace that automates sprint planning, project execution, and cross-tool collaboration for engineering teams using autonomous agents.
- Built features across a TypeScript / Next.js codebase with LLM agent integrations and deep third-party tool integrations; collaborated via pull requests, code review, and issue tracking.
Tailoring: for AI roles emphasize "autonomous agents / LLM integrations"; for FS roles emphasize "TypeScript, Next.js, production SaaS codebase"; quantify (users, PRs merged, features shipped) as soon as you can.
PROJECTS - SWAP ORDER (lead with the 2 most relevant; keep 3–4 total)
Order per track - AI: 1,2,4,3 · ML: 2,5,1,3 · FS: 1,6,7,4 · CV: 3,5,2,1 · GEN: 4,1,2,3
1. AI Book Assistant - Next.js, TypeScript, Vapi AI, Tailwind CSS · [Live] · Jun 2026
- Built a voice-enabled AI assistant that turns uploaded PDFs into interactive conversations, combining document parsing, RAG (Retrieval-Augmented Generation), and real-time speech.
- [AI variant] Implemented conversational voice pipeline (speech-to-text → LLM → text-to-speech) with streaming responses and context-aware document retrieval.
- [FS variant] Designed and deployed the full stack: Next.js App Router frontend, serverless API routes, and Vercel CI/CD with a responsive Tailwind CSS interface.
2. AI Customer Churn Prediction System - Python, XGBoost, Streamlit, Gemini AI, Plotly · [Live] · Feb 2026
- Developed an end-to-end telecom churn prediction application: data preprocessing, feature engineering, XGBoost model training, and an interactive Streamlit analytics dashboard.
- [ML variant] Delivered probability risk gauges, customer health scores, and model evaluation workflow with persisted preprocessing artifacts for reproducible inference.
- [AI variant] Integrated Google Gemini AI to generate personalized, AI-driven customer retention recommendations from model outputs.
3. Eye Blink Morse Code Detector - Python, OpenCV, MediaPipe, NumPy · Jul 2025
- Built a hands-free communication system converting intentional eye blinks into Morse code text using EAR (Eye Aspect Ratio)-based blink detection and MediaPipe Face Mesh landmarks in real time.
- [CV variant] Tuned real-time video processing for stable detection across lighting conditions; designed accessible input for users with motor impairments.
4. Software Systems & AI Project Suite - Python, C++, JavaScript, OOP, CRUD, Git · 2025–2026
- Unified collection of management systems, AI-powered games, and database applications built with OOP (Object-Oriented Programming) design and SDLC (Software Development Life Cycle) practices.
5. Pentagram Image Diffusion - TypeScript, Diffusion Models · Aug 2026
- Text-to-image generation app: iterative diffusion refines random noise into coherent images guided by natural-language prompts.
6. Ghoomora - Travel Platform - Next.js, TypeScript · [Live] · Jul 2026
- Built and deployed a northern-areas travel application with destination discovery and trip planning.
7. RetroVerse - Retro Gaming Hub - Next.js, TypeScript · [Live] · Jul 2026
- Built and deployed a retro gaming hub with a game library UI and responsive design.
8. Site Scrapper - Python, Selenium, Beautiful Soup · [Live] · Jul 2026
- Production web scraping tool for automated site data extraction with a deployed helper interface.
Include project 8 only for Automation/Scraping/Data-Engineering roles (replace 5–7).
TECHNICAL SKILLS - SWAP (reorder so the job's keywords appear in the FIRST line; fix: "Frameworks" not "Framworks")
Languages: Python, TypeScript, JavaScript (ES6+), SQL, C++, C, HTML5, CSS3
AI / ML
LLM (Large Language Models), RAG (Retrieval-Augmented Generation), Prompt Engineering, LangChain, LangGraph, Machine Learning, Computer Vision, OpenCV, MediaPipe, XGBoost, scikit-learn, Pandas, NumPy, OCR, Hugging Face, OpenAI API, Google Gemini, Anthropic Claude, Ollama, Groq
Frameworks & Libraries: Next.js, React.js, Node.js, Express.js, FastAPI, Flask, Tailwind CSS, Streamlit, Selenium, Beautiful Soup, Pydantic, Zustand, Framer Motion, GSAP, Plotly
Tools & Platforms: Git, GitHub, Docker, PostgreSQL, MongoDB, Supabase, Vercel, Postman, Linux, Figma, VS Code, Jupyter / Google Colab, Kaggle, Wireshark, Cisco Packet Tracer
Core CS: Data Structures & Algorithms, OOP (Object-Oriented Programming), SDLC, REST APIs, Computer Networks, Agile/Scrum
Per-track first line - AI: AI/ML block first · ML: AI/ML block first (lead with XGBoost, scikit-learn, Pandas) FS: Frameworks block first · CV: lead with OpenCV, MediaPipe, Computer Vision · GEN: Core CS first.
EDUCATION - CONSTANT
FAST NUCES University, Pakistan - Bachelor of Science in Computer Science · Aug 2022 – Jun 2026
GPA omitted intentionally. Add "Relevant coursework: ..." only if the posting lists specific topics you covered.
LANGUAGES - CONSTANT
English (Fluent) · Urdu (Native) · Arabic (Beginner)
PRE-SUBMIT CHECKLIST (delete before export)
- Job title in the posting appears verbatim in your summary
- Top 6 keywords from the job description appear in Skills AND at least one project bullet
- Every acronym expanded once: RAG, LLM, ML, OOP, OCR, SDLC
- Halyo experience line kept and angled to the role track
- Projects reordered per track; max 4 shown
- Exported as PDF named
Mohammad_Bilal_<Company>_<Role>.pdf