# about.md

Badal Jha

Senior Software Engineer

I'm a backend engineer who builds the systems behind AI features. I work on LLM gateways, agent platforms and the Java services that keep them fast, affordable and reliable.

## about README.md

I'm happiest deep in a system's internals, profiling, optimizing and scaling it until it's fast, lean and boring in production. My path went from building product features to distributed systems, and now to making AI practical at scale. I like clean design, measuring before guessing, and writing about what I learn along the way.

## now ~/.config/now.toml

building: = "Central LLM gateway and agent orchestration platform"
learning: = "distributed systems, agentic AI"
reading: = "Meditations by Marcus Aurelius"
updated: = "2026-10"

## writing ls -lt blog/

open blog/ →

## projects tree projects/ -L 1

open projects/ →

llm-gateway/ Vendor-agnostic gateway with one API for OpenAI, Gemini and Anthropic. Dynamic routing, fault tolerance and automatic fallback keep requests flowing when a provider fails, with no vendor lock-in. Powers 200+ GenAI use cases at Naukri.com. Java · Spring Boot · Kafka @ naukri.com · internal
llm-router/ Routes each request to the cheapest model that can handle it, using a fine-tuned classifier. Cut LLM costs by 35%. Python · DeBERTa · Java @ naukri.com · internal
agent-framework/ Multi-agent framework where agents keep context, use tools over MCP and hand work to each other to automate multi-step workflows. Java · MCP · LLMs @ naukri.com · internal
voice-sdk/ Real-time speech SDK (text-to-speech, speech-to-text, speech-to-speech) over WebSockets, with one interface for several voice providers. Java · Python · WebSockets · Kafka @ naukri.com · internal

## contact badal --hello

email: badalkjha@gmail.com
github: github.com/badal-jha
linkedin: linkedin.com/in/badaljha
phone: +91 98115 87430
rss: /rss.xml