Devatraj Bhattacharjee
Capability pages explored:0 of 5Email me

What I do, 3 of 5

Ship AI features people use

AI that sits inside your product's main flow, where it moves a metric, not a chatbot bolted on the side.

35%more conversions from one LLM feature

Who it's for: Product managers and founders who want AI to change a number, not just a demo.

The problem

Try it

A feature I shipped, rebuilt as a demo. Pick a suggestion and send it.

28, Software engineer

Suggested by AI Pick one

Illustration with a sample profile

Where I've done it

Jeevansathi

AI first messages

Built AI first-message suggestions. When someone sends an interest, an async Kafka worker reads both profiles and asks Qwen (on Vertex AI) for personalised openers, so the sender can pick one instead of sending a blank interest.

  • A/B tested on 50% of users

Jeevansathi

LLM-regenerated listings

Led LLM integration across a Kubernetes-deployed React/Redux and Spring Boot stack, using Structured Outputs (JSON Schema) to regenerate listing data.

  • 35% increase in conversion

Jeevansathi

Real-time GenAI matchmaking

Delivered a real-time GenAI matchmaking engine using Spring Boot, Openfire (XMPP) and Aerospike, orchestrating GPT, Gemini and Qwen through a LangChain/LlamaIndex RAG pipeline on RedisVL vector search.

  • 99.9% uptime

Kundli match

Ashtakoota, 8 kootas

Good match

AI conversation starter

You both have the Moon in earth signs. Ask about the last thing they built with their hands.

Illustration with sample scores

milanSathi. AI conversation starters grounded in both charts.

How I'd do it for you

  1. Pick the moment

    Find the step in your funnel where people stall.

  2. Ground it in your data

    Generate from the user's real context, with structured output your app can trust.

  3. A/B test it

    Ship to a share of users and keep it only if the metric moves.

Need this?

Tell me what you're working with. I'll tell you honestly whether it's worth doing and roughly what it takes.

Email me about this