At WAMO Labs, I deliver production-grade AI and full-stack solutions for client projects across e-commerce, automation, and knowledge assistants. I built RAG pipelines with LangChain and Pinecone that improved answer accuracy from 60% to 95% and cut lookup time by 60%.
I designed LangGraph agent workflows and custom MCP implementations to standardize how LLMs interact with application services. I also integrated Whisper for multilingual, near real-time transcription, reaching 95% accuracy at under two seconds of latency.
At WAMO Labs, I orchestrate asynchronous AI workflows with Temporal and RabbitMQ and deploy Docker services on AWS. I develop applications with React and TypeScript front ends and Python back ends using MongoDB, PostgreSQL, DynamoDB, and RDS.
At Turing, I evaluated model outputs for SFT and RLHF pipelines, reviewed LLM-generated code, and identified failure modes such as hallucinations and instruction-following errors. I also built RVio, an AI-powered assistant for querying SQL inventory data in natural language, and Listen, a real-time Pakistan Sign Language interpreter that converts signs into text and speech.
