---

● ● ●   shah-bakhsh-os ~ v2026.8 { "engineer": "Shah Bakhsh", "role": "AI Engineer & Autonomous Systems Architect", "location": "Pakistan", "email": "shahbakhshtech@gmail.com", "2026_impact": { "annual_contributions": "2,019+ Commits & Code Activity", "public_repositories": "11 Projects", "certifications": "6 Verified AI/ML Accreditations" }, "core_competencies": [ "Autonomous Multi-Agent Swarms & State Machine Orchestration (LangGraph, CrewAI)", "Enterprise RAG Pipelines & Vector DB Architectures (Pinecone, ChromaDB, FAISS)", "LLM Fine-Tuning (PEFT, LoRA, QLoRA) & Context Optimization", "Real-Time Computer Vision & Multimodal Deep Learning (OpenCV, PyTorch)", "Production MLOps Pipelines, FastAPI Serving & Docker Deployments" ], "current_research": "Model Context Protocol (MCP), Autonomous Self-Correction & Multi-Agent Swarms", "status": "🟢 Available for High-Impact AI Roles, Technical Advisory & Consulting" }

---

⚡ Performance & Contribution Dashboard

🚀 2026 Contribution Metrics Dimension Real-Time Status 🔥 Annual Contributions 2,019+ Commits ⚡ Open Source Projects 11 Repositories 🎓 Verified Certifications 6 Accreditations 🤖 Domain Focus Agentic AI & RAG 💼 Availability Open for Hiring & Projects

🎯 Engineering Highlights 🔹 Autonomous Swarms: Multi-agent workflows with LangGraph & CrewAI 🔹 RAG Architecture: High-precision vector search & re-ranking 🔹 Computer Vision: Live camera video stream face tracking & classification 🔹 MLOps Infrastructure: Fast REST APIs, Docker, and CI/CD pipelines

🐍 Contribution Snake Animation

---

🧠 Core AI Engineering Architecture

🤖 Agentic AI & Multi-Agent Frameworks Orchestration: LangGraph, CrewAI, AutoGen, LangChain. Capabilities: Multi-agent state machines, autonomous tool selection, persistent long-term memory, self-correction loops & human-in-the-loop validation. Observability & Tracing: LangSmith, DeepEval, Phoenix RAG evaluation.	🧠 LLMs & Advanced RAG Systems Architecture: Hybrid RAG, HyDE (Hypothetical Document Embeddings), Re-ranking, Context Compression & Semantic Chunking. Vector Databases: Pinecone, ChromaDB, FAISS, Qdrant, PostgreSQL (pgvector). Evaluation: RAGAS framework for precision, recall, and hallucination containment.
👁️ Computer Vision & Multimodal Deep Learning Stack: OpenCV, PyTorch, TensorFlow, MobileNetV2, CNNs, YOLO. Applications: Real-time face recognition attendance logging, mask detection access control, and live video stream analysis.	⚙️ MLOps, Cloud & Production Delivery Backend: FastAPI, Streamlit, Next.js, Node.js, Python. DevOps: Docker, GitHub Actions CI/CD pipelines, Git version control. Reliability: Data drift monitoring, REST microservices, low-latency inferencing & production deployments.

---

🛠️ Technical Stack & Frameworks

🚀 Programming Languages

🤖 AI, Machine Learning & Deep Learning

🧠 LLMs, GenAI & Multi-Agent Frameworks

💾 Vector Databases & Storage

📦 MLOps, Infrastructure & Web Tools

---

⭐ Featured Flagship Projects

🤖 Zirak AI LLMs · Agentic Workflows · LangChain Autonomous multi-step AI assistant with persistent memory, custom tool invocation, and multi-tier reasoning chains for complex tasks.	🎯 Balochi AI Agent NLP · Low-Resource Dialects · Python Specialized AI agent system designed for Balochi natural language processing, OCR document extraction, and dialect translation.
🧭 PathPilot AI TypeScript · Pathfinding · Autonomous AI Intelligent pathfinding and navigation agent leveraging graph search optimization and real-time decision models.	👁️ Face Recognition Attendance Computer Vision · OpenCV · Deep Learning Automated real-time attendance system using facial detection algorithms. Logs camera stream identities automatically.
🛡️ Zrah - Phishing Cyber Guard Transformers · HuggingFace · Cyber AI Transformer-based NLP system with a real-time phishing URL detection engine classifying malicious web URLs to block attacks.	😷 Face Mask Detector AI CNN · MobileNetV2 · TensorFlow High-precision real-time face mask classification system developed for access control & public safety compliance.

---

📜 Verified Professional Certifications

---

📫 Let's Connect