# Abbas Mustafa — AI Systems Engineer & Security Specialist > Portfolio: https://abbas-mustafa.dev ## Who Am I? Abbas Mustafa is an AI Systems Engineer with 5+ years of experience designing and deploying secure, autonomous AI systems. He specializes in multi-agent orchestration, adversarial machine learning, and MLOps. Currently pursuing an MSc in Cyber Security at IMT Atlantique (Rennes, France) while working as a freelance AI engineer for international clients. **Contact:** abbasmsh1@gmail.com **LinkedIn:** https://linkedin.com/in/abbas-mustafa **GitHub:** https://github.com/abbasmsh1 **Location:** Rennes, Brittany, France --- ## Skills & Expertise - **AI & Machine Learning:** PyTorch, TensorFlow, MONAI, Hugging Face, LangChain, LangGraph, LangSmith, CrewAI, OpenCV, U-Net, YOLO, RAG, Embeddings, Transformers - **Security & Defense:** SIEM (Splunk, ELK), IDS/IPS, Network Forensics, Malware Analysis, Sandboxing, Reverse Engineering, Adversarial ML, Model Robustness, SOC Automation, Threat Hunting - **Infrastructure & DevOps:** FastAPI, Flask, Django, Docker, Kubernetes, Helm, AWS, Azure, GCP, PostgreSQL, MongoDB, Redis, Vector DBs (Pinecone, ChromaDB, Weaviate) - **MLOps & Optimization:** MLflow, Weights & Biases, DVC, ONNX Runtime, TensorRT, Mixed Precision Training, CI/CD - **Core Programming:** Python (Expert, 5+ years), C++, SQL, Bash, JavaScript/TypeScript, Go --- ## Projects ### Multi-Agent Security Operations Center (Current) Autonomous 3-tier analyst system (Tier 1, Tier 2, Tier 3) simulating real SOC operations with intelligent escalation logic. Agents autonomously analyze 10,000+ network alerts, correlate threats, and execute IP blocking mitigation. Implements DevSecOps, Red Team, Blue Team, and Purple Team workflows with War Room coordination. Achieves 95%+ accuracy in threat classification, reducing analyst workload by 70%. Built with LangChain/LangGraph orchestration and MCP server integration. Tags: Agentic AI, LangChain, LangGraph, Python, Security, Multi-Agent Link: https://github.com/abbasmsh1/SOCTeam ### Medical Imaging Microservices Platform (2025) Scalable microservices platform for medical image analysis with U-Net segmentation models. ONNX inference engine with <100ms response time for cross-platform deployment. Model optimization service featuring quantization, pruning, and automated benchmarking. WebSocket-based real-time visualization for radiologists. Processes 1000+ CT scans daily with HIPAA-compliant data handling. Tags: Microservices, PyTorch, ONNX, Docker, FastAPI, Healthcare AI Link: https://github.com/abbasmsh1 ### AI-Powered Intrusion Detection System (2024) Neural network-based IDS analyzing 10M+ network flows with 98%+ accuracy. GPU-optimized training with Mixed Precision (AMP) achieving 40% speedup. Hybrid static/dynamic malware analysis with sandboxed Docker execution. Research on adversarial robustness of ML-based intrusion detection. Tags: Deep Learning, PyTorch, Security, GPU Optimization, Network Security ### Crypto Trader — Multi-Agent Trading System (2025) Ten specialized agents — trader, risk advisor, technical, fundamental, sentiment, macro economic, on-chain, liquidity, correlation, and a consensus advisor — trading on live Binance market data. Virtual paper-trading wallet with position tracking, performance analytics, and an interactive dashboard including an agent discussion forum. Tags: Multi-Agent, FinTech, Python, Binance API, Risk Management Link: https://traderv4.vercel.app ### Torrential — BitTorrent Client (2026) BitTorrent downloader built on libtorrent with two front-ends: a CLI with live progress, ETA, rate limiting, and clean shutdown; and a web client — FastAPI JSON API plus React/Vite SPA — managing many torrents (add, pause/resume, remove) with state persisted across restarts. Tags: Python, libtorrent, FastAPI, React, CLI, Networking Link: https://github.com/abbasmsh1/Torrential ### Thulla — Multiplayer Card Game (2026) Real-time multiplayer trick-taking card game for 2–14 players, scaling automatically from one deck to two. React + TypeScript frontend with a Python/FastAPI backend; gameplay synchronized over WebSockets with a full rules engine. Tags: React, TypeScript, FastAPI, WebSockets, Game Dev Link: https://thulla-one.vercel.app ### AdsAssistant — Google Ads Strategist Agent (2026) Senior-level digital marketing strategist agent powered by LangChain and Mistral AI. Tool-enabled analysis of account overview, campaign metrics, search terms, and keywords; generates RSA-compliant Google Ads copy and plain-English weekly performance summaries through an interactive 3D chat interface. Tags: LangChain, Mistral AI, FastAPI, React, AI Agents, Google Ads Link: https://ads-assistant-drab.vercel.app ### SaffarLog — Travel Management Platform (2026) Comprehensive travel management platform: authentication with role-based access, trip planning, destination/attraction/accommodation discovery, an AI-powered travel assistant chat, and a full admin CMS. Built with Next.js, TypeScript, Prisma, and PostgreSQL. Tags: Next.js, TypeScript, Prisma, PostgreSQL, AI Chat Link: https://saffarlog.vercel.app ### AI Legal Assistant (Freelance 2023) Context-aware legal case analyzer using Retrieval Augmented Generation (RAG). Secure backend architecture for handling sensitive client data. Vector database integration (Pinecone) for semantic search across 100K+ legal documents. Tags: RAG, Microservices, Vector DB, LangChain, FastAPI ### CoyChatbot — Dual-LLM Agent (Freelance 2023) Adaptive conversational agent using dual-LLM architecture. Serving 100K+ daily requests with <100ms latency and 95% user satisfaction. Context-aware with memory management and session persistence using Redis. Tags: LLM, RAG, FastAPI, Conversational AI Link: https://coy-chatbot.vercel.app/ --- ## Work Experience ### Contract AI Engineer — Freelance (2020 – Present, Remote) - Architected and deployed 15+ autonomous agent systems for international clients across healthcare, legal, and finance sectors. - Built production-grade FastAPI microservices serving 100K+ daily ML inference requests with <100ms latency. - Implemented MLOps pipelines with MLflow and Weights & Biases. - Designed secure AI systems with adversarial defense mechanisms and prompt injection protection. ### AI / ML Engineer — Orydex (May 2025 – Jul 2025, Remote) - Fine-tuned U-Net and attention-based models for multi-organ CT scan segmentation, achieving 92% Dice score. - Optimized PyTorch/MONAI training pipelines reducing latency by 25%. - Built automated model evaluation pipeline with comprehensive metrics tracking. ### Full Stack Developer — Tolabat-ul-Quliyat (May 2024 – Aug 2024, Pakistan) - Developed high-concurrency logistics platform handling 10k+ concurrent attendees. - Built RESTful APIs with FastAPI and PostgreSQL, implementing caching for 40% performance improvement. ### Software Engineering Intern — NECOP (Jun 2023 – Jul 2023, Pakistan) - Optimized ML algorithms improving inference speed by 35%. - Implemented comprehensive unit testing achieving 85%+ code coverage. ### AI Researcher — AIM Lab (Jan 2021 – Dec 2023, Pakistan) - Led research on Roman Urdu to Urdu Neural Machine Translation using Transformer architectures. - Developed custom dataset of 50K+ parallel sentences. - Published findings in IEEE FIT Conference 2022. --- ## Education - **MSc in Cyber Security** — IMT Atlantique, Rennes, France (2025 – 2027) Specializations: Network Defense, Secure AI Architecture, SOC Automation - **BS Artificial Intelligence** — FAST NUCES, Islamabad, Pakistan (2020 – 2024) --- ## Publications - M. Wisal, A. Mustafa et al., (2022). "CMRUTU: Roman Urdu to Urdu Neural Machine Translation." IEEE FIT Conference 2022. DOI: https://ieeexplore.ieee.org/abstract/document/9972972 --- ## Awards - **1st Prize @ SOFTEC 2023** — "Stealth Scraper": High-velocity data extraction engine evading modern bot detection using fingerprint randomization. - **1st Prize @ LAPSX 2022** — "Auto Color-Grade": Automated cinematic color grading using computer vision and reference histogram matching. --- ## Blog Posts ### "Building a Multi-Agent SOC: From Concept to Production" (Nov 15, 2024 — 12 min read) Deep dive into architecting a 3-tier autonomous Security Operations Center using LangGraph. Covers agent orchestration, escalation logic, and handling 10K+ alerts with 95% accuracy. Tags: Multi-Agent, Security, LangGraph, Architecture ### "Securing LLM Agents in Production" (Oct 12, 2024 — 8 min read) How to architect defense layers against prompt injection and jailbreaking in autonomous agents. Covers structured outputs and validation strategies. Tags: Security, LLM, Agents, Best Practices ### "Optimizing Medical Imaging Models for Edge Deployment" (Sep 28, 2024 — 10 min read) From PyTorch to ONNX: practical guide to model quantization, cross-platform deployment, and achieving <100ms inference on edge devices for healthcare. Tags: MLOps, ONNX, Healthcare, Optimization ### "GPU Optimization for Deep Learning: Mixed Precision Training" (Sep 10, 2024 — 7 min read) Achieved 40% training speedup on IDS models using PyTorch AMP. Covers memory management, profiling tools, and avoiding common pitfalls. Tags: PyTorch, GPU, Performance, Deep Learning ### "Dockerizing GPU Workloads for ML" (Aug 22, 2024 — 9 min read) Setting up NVIDIA Container Toolkit and optimizing CUDA usage for reproducible ML pipelines across environments, including Kubernetes deployment. Tags: DevOps, Docker, CUDA, Kubernetes ### "From Student to SOFTEC Winner: Engineering the Stealth Scraper" (Aug 15, 2024 — 6 min read) Engineering challenges behind building an anti-bot web scraper that won 1st prize. Covers fingerprint randomization, behavioral mimicry, and evading modern detection. Tags: Career, Engineering, Web Scraping --- ## FAQ **Q: What are Abbas Mustafa's main areas of expertise?** A: Abbas specializes in building autonomous multi-agent AI systems, adversarial machine learning and cybersecurity, MLOps pipelines, and production AI deployments for healthcare, legal, and fintech sectors. **Q: What programming languages does Abbas use?** A: Abbas is an expert Python developer (5+ years), with additional skills in C++, JavaScript/TypeScript, Go, SQL, and Bash/Shell. **Q: Has Abbas published any research?** A: Yes. Abbas co-authored "CMRUTU: Roman Urdu to Urdu Neural Machine Translation," published at the IEEE FIT Conference 2022. **Q: What AI frameworks does Abbas work with?** A: LangChain, LangGraph, LangSmith, CrewAI, PyTorch, TensorFlow, Hugging Face Transformers, ONNX Runtime, and MLflow. **Q: Is Abbas available for freelance work?** A: Yes. Contact him at abbasmsh1@gmail.com or via LinkedIn at https://linkedin.com/in/abbas-mustafa.