I build AI systems
from prototype
to production.
>~/projects $ ls -la
4 entries
./01
AgenticEDA
Turn raw datasets into interactive notebooks, narrative reports, and explainable findings — autonomously through simultaneous agents.
↗
./02
Monty
Agentic system that converts classroom observations into an anonymized behavioral knowledge graph and research-backed recommendations.
↗
./03
JurasSigLIP
A text-guided segmentation prototype fusing DINOv3 vision tokens with SigLIP2 text tokens via cross-attention to produce precise segmentation masks.
↗
./04
Recipe Recommender
AI tool surfacing variations among 230k recipes based on ingredient availability, with LLM-tuned visualizations and a semantic recommendation engine.
↗
>
~/about $ cat README.md
Make it intelligent.
Make it robust.
Make it real.
I'm an ML Engineer building applied AI systems with a focus on reliability, evaluation, and product usefulness. My work spans agents, computer vision, and ML infrastructure — with an eye toward models that survive contact with real users. When I'm not building, I'm usually watching Pokémon VGC or playing tennis.
Based in San Francisco · Open to Relocate
>
indro@remote:~$ open --connections
$
GitHub
github.com/indrayudd
↗
$
LinkedIn
linkedin.com/in/indrayudd
↗
$
Résumé
drive.google.com · PDF
↗
>
~/connect $ ping indrayudd1@gmail.com
Say hello
Available for Work · San Francisco · 2026