AI and Enterprise Architecture with Rahul Dharankar at DeveloperWeek 2026

In this conversation from @devnetwork4615's DeveloperWeek 2026 at the San Jose McEnery Convention Center, Dave Vanz sits down with Rahul Dharankar, Head of AI at MarcoPolo, to explore what really happens when artificial intelligence moves from hype to production inside enterprise environments. Rahul breaks down why most companies are not struggling with prompts, but with production. While large language models perform well with public information, the real value inside organizations lives in private systems like data warehouses, CRMs, ERP platforms and internal knowledge bases. The challenge is not intelligence. It is integration, governance and control. The discussion dives into: • Why enterprise AI adoption stalls after pilot programs • Security and compliance concerns around internal data access • Cost control and token management at scale • Ownership versus dependency in a SaaS-driven world • How intent-based controls and controlled execution layers reduce risk • Why query-driven AI is more reliable than probabilistic text generation alone The role of global engineering teams in scaling AI infrastructure Rahul also shares his perspective as a multi-time startup founder and operator who has built and exited technology companies, worked in open source ecosystems and now focuses on solving the architectural friction that prevents AI from embedding into real workflows. This interview is for CTOs, engineering leaders and technical executives thinking about how to move from AI experimentation to enterprise grade execution. If you are asking whether AI actually works inside real companies, this conversation explains what must be in place for it to work responsibly, securely and at scale.