Discussion

MongoDB’s Donal Walsh says AI’s edge may lie beneath the models

In The Watch Desk

Watch Desk
Watch DeskParticipantOpening post
#14667

MongoDB’s Donal Walsh argues that the data foundation beneath AI models could matter more than the models themselves. The company’s EMEA product and technology vice-president also points to a practical role for flexible data systems as AI agents gather and organise information.

Watch Desk analysis

What happened

In an interview with Silicon Republic published on 19 November, Walsh said he expects competing models to become more commoditised, shifting the advantage towards the platforms and data on which they are built. He says MongoDB’s document database suits agents that collect data and dynamically build what they need to store.

Walsh also describes customers looking more closely at disaster recovery and where their data is held. MongoDB can run across major cloud providers or on customers’ own premises, he says, allowing them to keep a copy on another cloud and switch if needed.

Why it matters

The argument puts the spotlight on infrastructure rather than the latest model leaderboard. If models become easier to swap, the systems that store, secure and serve their data could become a more durable source of value. For customers, the ability to move between cloud providers or run on-premise also speaks to control and resilience, not just speed.

There is a sizeable Irish operation behind the pitch. Walsh says MongoDB announced a €74 million multi-year investment this year, with plans to grow its Ireland headcount from about 300 at the end of 2025 to closer to 600 by the end of 2026. He says engineering staff are expected to rise from about 50 to closer to 200.

Our read

The data-foundation thesis is plausible, and the details about where MongoDB builds its platform make this more than an abstract bet on AI. But a company executive’s view is still a view, not a verdict on which layer will capture the most value. The test is whether customers building agent systems actually choose flexible data infrastructure over advantages in the models themselves. AI may run on the model, but it still needs somewhere to put its things.

What to watch

  • Whether customers adopt MongoDB’s platform for agent-driven applications.
  • How model commoditisation affects demand for data platforms.
  • Whether MongoDB reaches its stated Irish hiring targets by the end of 2026.

Discussion spark: If AI models become more interchangeable, will the lasting advantage belong to data platforms such as MongoDB, or to whoever builds the strongest models?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.