Shreyash Bante
I build production AI for the physical world — agents, RAG, and ML pipelines on Databricks.
I'm a Data Engineer at Cummins, building production ML and AI agent workflows for engine diagnostics on Databricks. I own a unified pipeline that productionizes 25+ ML and rule-based diagnostic models, generating real-time alerts for technicians — work that has driven $1.2M+ in savings for the Engine Business Unit.
Right now I'm going deeper into agentic AI for industrial use cases — multi-agent systems on Databricks combining RAG over technician notes, natural-language data queries, and fault-code retrieval.
Previously: Azure data engineering at Motifworks and Accenture.
Writing publicly about the journey from data pipelines into industrial AI — what works on Databricks, what breaks, what I learn shipping ML in heavy industry. DMs open.