At Ingenero Technologies Inc, I architect and develop real-time, constraint-aware AI decision and optimization systems that help plant operators meet run-length and yield targets. The platform serves multiple production sites and contributes to more than $1M in combined annualized business impact.
I also engineered MLOps and CI/CD workflows that reduced deployment lead time by about 40% while maintaining about 99.9% uptime. My work includes an in-process ML training engine and failure-risk and RUL-prediction pipelines that improved predictive accuracy by about 70% over classical baselines.
I built TherapyBro, an AI-native mental wellness platform for LLM-driven chat and peer support, using FastAPI, LangGraph, and a Neo4j knowledge graph. It serves 1,000 monthly active users with infrastructure costs below $100 per month, and includes a crisis-keyword safety override with 100% recall.
At Aditya Birla Group, I designed a distributed ML framework on HPC clusters that reduced training time from 10 hours to 3, and delivered 98% accuracy in sentiment analysis by fine-tuning BERT, BART, and XLNet. I work across model development, evaluation, MLOps, and full-stack delivery.
