asif shah

At IKONIC DEV, I integrate LLM-based models into FastAPI production backends, owning prompt tuning, inference pipelines, and live AI features.

I design retrieval and RAG data pipelines that connect AI services with internal data sources for contextual responses. I also wire APIs and third-party integrations into downstream product systems while monitoring performance and reducing latency.

Previously at Devtrest, I built RAG pipelines for website-based AI products and shipped agentic workflows that connected agents to product actions and APIs. I supported end-to-end delivery, from prompt and inference tuning through integration, release, and webhooks.

My projects include RAG-powered conversational AI, a real-time AI call system using Twilio and ElevenLabs, and a multi-agent research system for retrieval, reasoning, and knowledge synthesis. I build with Python, FastAPI, React, OpenAI API, LangChain, LangGraph, Pinecone, FAISS, and SQL databases.