Priyadarshi Lahiri

Fourteen years building AI that reaches production — from knowledge graphs and predictive machine learning in 2012, through cloud-native data platforms, to today's LLMs, agents, and retrieval-augmented generation. I've twice been the first CTO in a company: building engineering organizations from nothing to 46 people, taking two companies from seed through Series A, and putting machine learning into production at HSBC, HCL, and Wipro under ISO and SOC2 compliance. The years since have been spent inside market leaders — MongoDB, Miro, and Neo4j — where the hard part is scale: large installed bases, reliability commitments, and the operational discipline both demand. Product has run alongside engineering throughout my career, from roadmap and customer ownership in the founding CTO roles to owning MongoDB's full Kubernetes operator portfolio, Community through Atlas. My current title as a Technical Product Manager formalises that, rather than starting it. The technical work hasn't stopped either. Distributed system design and solution architecture for enterprise deployments stay at the centre of what I do — at Neo4j, that means clustering, backup, and disaster recovery, plus the graph and vector infrastructure customers run their AI on. I recently built a production-ready Kubernetes operator end-to-end using Claude Code, working out first-hand where agentic AI genuinely speeds delivery and where its output still needs careful review.