Who is Munna Suprathik?

Munna Suprathik is a Generative AI Engineer based in Bengaluru, India. He builds production RAG systems, multi-agent pipelines, and LLM-powered apps, and works across Claude, GPT, and Gemini.

What does Munna Suprathik do?

He builds RAG systems, AI agents and chatbots, LLM-powered web apps, AI integrations into existing products, and Shopify stores. His deeper focus is agentic systems and automation with LangGraph, MCP, and A2A.

Is Munna Suprathik available for hire?

Yes. He takes on freelance and contract work. The fastest way to reach him is email at hello@munnasuprathik.com or LinkedIn.

What has Munna Suprathik built?

Live products include Tend, an agentic task app; Margam AI, a RAG study-abroad platform; Dissect, a viral-content analysis SaaS; and Quiccle, an AI discovery hub.

What is Munna Suprathik's tech stack?

Python, LangChain, LlamaIndex, LangGraph, vector search, Pinecone, Cohere Rerank, HyDE, parent-child chunking, LLM-as-judge evaluation, function calling, MCP, A2A, and LoRA and PEFT fine-tuning. He builds across Claude, GPT, and Gemini.

Munna Suprathik

Notes

Short, honest notes on building with AI.

The RAG That Retrieved the Wrong Thing — A production RAG system answered confidently with the wrong chunk. Why vector search finds similar, not right, and how reranking, HyDE, and parent-child chunking fixed it.
When I Trusted the LLM to Grade Itself — LLM-as-judge handed 9 out of 10 to answers that were flatly wrong. How to make an LLM judge trustworthy with a rubric, span quoting, and a human answer key.
The Agent That Looped Forever — A multi-agent LangGraph pipeline burned tokens going nowhere, with nothing in the logs. How step budgets, action memory, and circuit breakers stop agent loops.
Fine-Tuning Was the Wrong Answer — A weekend fine-tuning a Mistral 7B with LoRA lost to a better prompt and a retrieval step. When fine-tuning actually helps, and when prompt plus RAG wins.
The Prompt That Worked in Dev and Died in Prod — A prompt passed every test I wrote, then broke on real user input. Why prompts are brittle, and how a set of real messy inputs catches it before you ship.
Cutting a Feature From Tend — A feature I was proud of made Tend worse. Why more options meant less doing, and how cutting it made the product better. A note on product decisions.