— Carl Sagan, 1994
An AI developer and relentless builder based in Bangalore. I spend my days teaching machines to think and my nights building things that probably shouldn't exist yet.
I've tried to build startups (and failed spectacularly), freelanced with clients across the globe, and somewhere along the way fell in love with the art of making intelligent systems that actually work.
I believe the best technology disappears into the background. It doesn't demand attention — it earns trust.
I build intelligent systems that learn, adapt, and scale. From fine-tuning LLMs to deploying end-to-end ML pipelines, I work across the full spectrum of AI and software.
An end-to-end AI backbone powering Email, Meeting, and Contract agents using RAG, CoVe-based self-consistency ensembles, and chain-of-thought prompting. Built scalable NLP pipelines with semantic fingerprinting and vector embeddings.
UNLEASH THE AGENTS ↗
Proving LLMs as MESA Optimizers. Designed controlled experiments using Llama-3.1-8B-Instruct, identifying emergent optimization patterns in attention heads. Analyzed implications for AI alignment and safety.
READ THE RESEARCH ↗
A next-generation AI-powered development tool that automates code review and suggests architectural improvements using deep learning models trained on millions of code repositories.
VIEW PROJECT ↗
A real-time analytics dashboard for monitoring ML model performance in production environments. Tracks drift, latency, and accuracy metrics with beautiful visualizations.
LIVE DEMO ↗
Built AI agents for automated educational content. Knowledge graphs for adaptive learning. Reduced teachers' content creation time by 70%. Engineered a pipeline curating 50,000+ scientific diagrams.
Applied AI to energy, manufacturing, and mining sectors. Fine-tuned detection models with 5–10% accuracy gains. Delivered end-to-end ML pipelines for 3+ industrial clients.
Bangalore, India. Where the obsession began.
Building AI solutions for clients across the globe. From MVPs to production systems. Every project a new universe to explore.
Tried. Failed. Learned. Building again. The loop never ends.