🔊CEOInterviews

Chirantan Desai on AI coding agents

From Agents in the Enterprise with MongoDB | Interrupt 26 · · LangChain and MongoDB

“Last week, 70% of the code that was checked in was written by a coding assistant. Seventy percent. OK, that's great. It also kind of scares me a little bit — because then I'm like, 'OK, what is the token bill?' And it's pretty high.”

Chirantan Desai
President, Chief Executive Officer & Director, MongoDB Inc.
AI coding agentsengineering productivitycost management

On , Chirantan Desai, President, Chief Executive Officer & Director at MongoDB Inc., spoke about AI coding agents during Agents in the Enterprise with MongoDB | Interrupt 26 on LangChain and MongoDB.

Agents in the Enterprise with MongoDB | Interrupt 26
Watch on YouTube at 20:35
Agents in the Enterprise with MongoDB | Interrupt 26
LangChain and MongoDB
Watch on YouTube at 20:35
Harrison Chase sits down with CJ Desai, CEO of MongoDB, for a fireside chat at Interrupt 2026, the agent conference by LangChain. Watch their full session for a deep dive on what's actually working (and not) for enterprises building AI agents at scale, why 2025's "year of agents" prediction didn't quite pan out, and so much more. 00:00 Introduction & welcome 01:01 MongoDB's founding principles & how they apply to AI 02:47 AI is all about unstructured data - how MongoDB fits in 03:16 Adding native search, vector search & embeddings 03:41 How different customer types use MongoDB for AI 04:43 Frontier Labs: research, inference & long-term memory 05:22 AI-native companies: the ElevenLabs case study 06:16 Enterprise customers: LangChain orchestration & internal agents 07:06 The three-legged stool: LLM + data + harness layer 08:01 The LangChain x MongoDB integration story 09:37 Open ecosystem vs. hyperscaler lock-in 10:36 Adobe's agent story: making the LLM the bottleneck 11:00 Will startup, enterprise & lab use cases converge? 11:29 Three buckets of agents: employee, partner & customer-facing 12:22 Why customer-facing agents are the real prize 13:00 A large insurance company's agent journey 14:26 The real blockers: stack choice, regulations & evals 15:44 Agent stack paralysis & why 2026 feels more stable than 2025 16:46 Token costs & the ROI of coding agents 18:06 MongoDB's internal stats: 70% of code written by AI 19:37 Token budgets & the human vs. tool cost equation 21:37 Open memory & the Context Hub announcement 22:07 System of intelligence: why context & memory will get bigger 23:37 Why real-time agents can't be built on data warehouses 24:47 Headless APIs & the future of software-via-UI 25:17 CJ's SaaS procurement strategy: cutting vendors in half 27:51 Which SaaS companies are safe vs. at risk 29:08 What makes a strong moat for Gen AI-native startups 29:56 Platform mindset: when 1 + 1 is more than 2 31:39 Wrap-up & thanks Extra resources: • Everything we shipped at Interrupt: https://www.langchain.com/blog/interr... • Meet LangSmith Engine: https://www.langchain.com/blog/introd... • About LangChain: https://www.langchain.com/
Chirantan Desai

About Chirantan Desai

President, Chief Executive Officer & Director · MongoDB Inc.

Chirantan "CJ" Desai, who became CEO of MongoDB six months ago, has been speaking at several events about the company's role in the AI era. At the Interrupt 2026 conference, Desai said that large enterprises' prediction that 2025 would be the year to create agents at scale "didn't pan out to be true," though he noted that technologies around observability and harnesses are maturing in 2026. He described coding assistants as the current "killer use case" for AI in enterprises. Desai also stated that MongoDB is only signing one-year contracts, citing uncertainty around hiring needs due to AI productivity gains, and said that 70% of code checked in at his company was written by a coding assistant. At MongoDB.local London 2026, Desai argued that the data layer "will not be commoditized" and has been "the only constant" through technological transformations. He described MongoDB as a unified platform integrating text search, vector search, and transactional data for building agents. Desai expressed skepticism about seat-based pricing models, advocating for consumption-based or outcome-based pricing instead. He also discussed the importance of multi-cloud resiliency, noting that most hyperscalers experienced major outages in 2025. Desai said that companies must either use AI to create new products or reinvent themselves on AI-first principles, and that simply rebranding as an AI company "does not work."

Profile compiled from Chirantan Desai's verified public interviews and appearances. See all quotes & transcripts →

More from Chirantan Desai MongoDB Inc. (MDB) Full Transcript Explore All Executives