MongoDB (MDB) Investor Day, September 2026

Dev, strategy

Dev back as interim CEO. Operational database + search + retrieval + agent memory. More AI-built apps should mean more workloads. Need to see that in consumption.

Voyage for embeddings and reranking. Atlas Agent Engine for memory and governance. Separate what is shipping from what is contributing revenue.

Atlas Agent Engine

Public preview, available to existing Atlas customers. Persistent short- and long-term memory, security guardrails, and flexibility across models and frameworks. Management expects it to drive additional Atlas consumption.

Mgmt expects GA late FY27 / early FY28, then traction through FY28. Watch production usage and incremental Atlas consumption.

Core products and retrieval

9.0, Atlas Infinite, self-managed search. Search + Vector Search for Enterprise Advanced went GA in June. Infinite still in public preview; mgmt does not expect meaningful revenue until FY29.

Vector Search customer count grew 4x over two years. AI-native customers account for roughly 3-5% of Atlas revenue, growing 165% year over year. Fast growth, still a small base.

Mike, three-year framework

• Total revenue growth: 20%+
• Atlas revenue growth: mid-20s, raised from 20%+
• EA and other: 10%+ ARR growth
• Non-GAAP operating margin: 100-200 bps of annual expansion
• Free cash flow conversion: 80-100%

The horizon narrows from three-to-five years to three. Margin expansion and cash conversion targets are unchanged. EA is framed in ARR to reduce the noise from multiyear revenue recognition.

Follow-ups

Rebuild the Atlas schedule around the mid-20s target. Separate existing workload consumption, new applications, and incremental AI usage. Keep management's target alongside our own assumptions.

Track Agent Engine adoption and the consumption it produces. Keep its preview-to-GA path separate from Atlas Infinite's longer revenue timeline. Test the margin bridge against continued investment in product and go-to-market.