Abdul Mueed Khan's Work | ContraWork by Abdul Mueed KhanAbdul Mueed Khan
AI & ML Engineer | Data-Driven & Scalable Apps
Air Quality Prediction System Multi-Agent Compliance Research & Action Assistant REACH AI is an end-to-end intelligent recruitment platform engineered to automate candidate screening, technical evaluation, and interview workflows using autonomous multi-stage AI agents.
Key Architecture & Features
Multi-Stage Agentic Evaluation: Orchestrated specialized LLM agents to conduct dynamic, context-aware technical interviews and automated resume parsing against role requirements.
Live Code Sandboxing: Designed isolated Docker execution environments to securely compile, run, and evaluate candidate code submissions in real time.
Real-Time Communication: Integrated low-latency WebSockets to handle seamless bidirectional audio, text, and state synchronization between the candidate and the AI interviewer.
Scalable Backend: Architected a modular FastAPI backend with PostgreSQL for structured candidate scoring, metrics tracking, and interview analytics.
Tech Stack
Python, FastAPI, LangChain/LangGraph, WebSockets, Docker, PostgreSQL, React.