ML Engineer
Embeddings, clustering, and the PEA pipeline. You should be comfortable with multilingual NLP, UMAP/HDBSCAN-class methods, and reasoning about evidence integrity — not just model accuracy.
CAREERS / DEHRADUN
Small team, Dehradun, high standards.
Embeddings, clustering, and the PEA pipeline. You should be comfortable with multilingual NLP, UMAP/HDBSCAN-class methods, and reasoning about evidence integrity — not just model accuracy.
FastAPI/Celery backend and React 18 console, packaged for air-gapped Docker/Kubernetes deployment. You care about systems that run where there is no internet to fall back on.
Pre-revenue, bootstrapped, and deliberate. We build for environments where failure is not a rollback — it is a consequence.