An agent can finish the demonstration before the engineering team has decided how the system should remember, recover, verify and stop.
AI Dev New York spends two days inside that unfinished architecture. Andrew Ng and DeepLearning.AI are bringing developers to New York on November 30 and December 1 for technical sessions on production agents, evaluation-driven development, memory, context and AI-assisted software engineering.
Andrew Ng, Yann LeCun, Salil Sethi, Saahil Jain, Joel Niklaus, Tariq Shaukat and Stephen Chin are among more than 35 speakers currently listed across the program.
The conference begins with Ng's AI Engineering Skills Map, followed by sessions on agentic workflows, evaluation, the data layer and product-minded engineering. Day two moves into production applications built with coding agents, more autonomous development systems, OS control, voice and research.
That sequence matters. The model may be the most visible part of the product, but the engineering record lives in the routing, state, permissions, observability and recovery work surrounding it. A system becomes useful in production when a team can explain what happened after the first plan changed.
Use the official page for current pricing, schedule, speakers and venue details.
2026-11-30 through 2026-12-01; 8:00 AM-6:00 PM ET daily.
Private New York venue, New York.