Kaleem Ahmed — AI, Backend & Full-Stack Engineer

Backend systems, end to end. I am a backend-heavy full-stack engineer based in Noida, Uttar Pradesh, India, currently at Astrea IT Services. I work mostly in Node, TypeScript, Python and FastAPI.

On the AI side I build retrieval-augmented generation over MongoDB Atlas Vector Search with Jina embeddings, running Gemini and Groq models: six AI surfaces on one shared retrieval substrate, with prompt and context isolation between them, streaming responses, memory consolidation and safety guardrails.

I build microservices, RabbitMQ event pipelines, payment and order flows, role-based authorization models, and the observability around them. Most of my depth is in the backend: where service boundaries go, where data should live, what happens asynchronously, and how a system behaves when one part of it fails.

Every project below has a written case study rather than a screenshot. They cover the architecture, the trade-offs I would defend in an interview, and the things I got wrong and had to fix.

Case studies

Elsewhere