Aaqil I.

I build RAG pipelines and LLM-powered applications that actually ship. My recent work includes a production-grade hybrid retrieval system (BM25 + FAISS + Ollama) deployed for a logistics company, significantly reducing their document lookup time across a large knowledge base. Here's what I typically build for clients: — Multi-agent systems using LangGraph and LangChain, handling complex, multi-step reasoning and task orchestration — RAG pipelines with hybrid retrieval (dense + sparse), local LLMs via Ollama, and vector stores like ChromaDB and FAISS — Computer Vision applications for object detection, classification, and domain-specific analysis — AI-powered web apps with React or Next.js frontends and Flask/Python backends — Workflow automation connecting LLMs to external APIs, tools, and data sources via N8N or custom integration I care about systems that are maintainable, explainable, and ready for real users — not just ones that pass a demo. 100% Job Success · Top Rated on Upwork BSc Computer Science — PUCIT CS50's Introduction to AI with Python (HarvardX) · Python for Data Science (Udemy)