WillItSurvive

Enterprise platform letting big companies build and govern AI and data pipelines in one place.

Hard to displace
Assessed 21 July 2026 · Evidence as of 21 July 2026
Strong structural protection; the business model survives ai pressure.
On the recordSince 21 July 2026 · Checkable from H1 2026 · Every call on the record Checkpoint: Planned US IPO
The call

Dataiku sells the neutral ground above Databricks and Snowflake, and its survival depends on enterprises still wanting one governed layer instead of picking a single cloud winner.

Dataiku builds tools that let large companies organize and run their data work across different cloud providers and software suppliers, while keeping track of where data comes from and making sure it follows rules like European AI laws. It is hard to copy because Databricks and Snowflake, the main rivals, are locked into their own products and cannot easily build a neutral middle layer without hurting their own business, and because Dataiku has deep relationships with 750+ large customers paying hundreds of thousands of dollars per year. The big question is whether Databricks will start winning new customers by bundling its own orchestration and governance tools together, or whether enterprises will decide to abandon the multi-vendor approach and commit entirely to a single platform.

Why
01
AI pressure: OpenAI, Anthropic, Snowflake Cortex. Foundation models can auto-generate simple ML workflows, pushing Dataiku to sell agent orchestration and governance instead of raw automation. Could AI do the core job? Only partly.
02
Bigger players: Databricks. Databricks grew to $3B ARR at 60% YoY with Mosaic AI Agent Framework and Unity Catalog, aiming to replace neutral layers like Dataiku's. Could a giant copy it? Probably.
03
What protects it. 750+ large enterprise customers pay $400K-500K a year for governance and multi-cloud neutrality that Databricks and Snowflake can't offer without cannibalizing their own lock-in.
What would change this call
Dataiku IPO filed or priced in 2026. Confirms public-market validation of its $6-9B speculative valuation range.
Named enterprise account publicly switches from Dataiku to Databricks or Snowflake Cortex. Would show the governance/orchestration moat is being breached in practice.
Dataiku's next ARR update shows growth below 15% YoY while Databricks stays above 40%. Would indicate Dataiku is losing relative growth momentum to the named incumbent.
Key numbers
  • $350M ARR (Oct 2025)
  • 750+ enterprise customers (Oct 2025)
  • 1,636 employees (Jun 2026)
Named risks
  • Core AutoML/no-code pipeline value being commoditized by frontier LLM agent builders. Severity: high
  • IPO timing uncertainty could delay or reprice below speculative $6-9B range. Severity: medium
  • Databricks and Snowflake shipping adjacent governance and agent features that could erode neutral-layer positioning. Severity: medium
  • Growth deceleration from prior multi-year doubling pace to about 17% ARR growth in nine months to Oct 2025. Severity: medium
Rivals, as read in July 2026
RivalDirectionEvidence
Databricksgaining onDatabricks ~$3B ARR growing ~60% YoY vs Dataiku $350M ARR as of Oct 2025, but treats Dataiku as partner not rival (Dec 2025)
SnowflakecontestingNamed Dataiku's Partner of the Year 5 years running through June 2026, while shipping own Cortex AI features
DataRobotgaining onDataRobot ~$1.5B valuation, ~$140M ARR Q3 2023, reportedly exploring strategic sale per Sacra Oct 2025
AlteryxcontestingAlteryx ~$970M revenue 2023, added Copilot/GenAI features December 2025
Questions people ask
Is Dataiku profitable in 2026?
Not confirmed. Dataiku stated in January 2025 that it was 'steadily slashing burn rate' with a 'near-term path to profitability,' but exact profit figures remain undisclosed. As of mid 2026 no public confirmation of profitability has been reported.
Is Dataiku planning an IPO?
Yes. Reuters reported October 1, 2025 that Dataiku hired Morgan Stanley and Citigroup to prepare for a possible U.S. IPO as early as H1 2026. Speculative valuation chatter cites $6B to $9B, but no filing or pricing had been confirmed as of mid 2026.
Could AI replace Dataiku?
Partially. Dataiku's original no-code AutoML pipeline building is exposed to commoditization by LLM agent builders, prompting its 2026 pivot to agent orchestration and governance. Its enterprise lineage, audit tooling for EU AI Act compliance, and multi-cloud lock-in across 750+ customers are harder for a model alone to replace.
Who is Dataiku's biggest competitor?
Databricks is the closest large-scale threat, with roughly $3B ARR and about 60% YoY growth versus Dataiku's $350M ARR as of October 17, 2025. Direct product rivals include DataRobot and Alteryx, while Snowflake remains a close technology partner rather than a rival as of December 2025.
Is Dataiku growing or slowing down?
Dataiku surpassed $300M ARR in January 2025 and $350M ARR by October 17, 2025, roughly 17% growth over about nine months, down from its earlier multi-year pace of more than doubling ARR in three years. It serves 750-plus enterprise customers as of October 2025.
How many people work at Dataiku?
Dataiku's own release states over 1,250 employees across 13 offices as of October 17, 2025. Third-party tracker Revelio Labs estimated about 1,450 employees by December 31, 2025, and Crustdata put headcount at 1,636 by June 30, 2026.
Timeline
  • 2026-06-24Launched NVIDIA-validated Manufacturing AI Blueprint with Kiji Inspector transparency framework
  • 2026-06-22Named Snowflake AI Data Cloud Partner of the Year for 5th consecutive year
  • 2026-03-09Rebranded as 'Platform for AI Success,' launched Cobuild and Reasoning Systems
  • 2025-10-17Announced surpassing $350M ARR with 750+ customers and 1,250+ employees
  • 2025-10-01Hired Morgan Stanley and Citigroup for potential US IPO as early as H1 2026
  • 2025-04-23Launched 'AI Agents with Dataiku' with agent orchestration and Cost Guard
  • 2025-01-16Announced surpassing $300M ARR milestone
  • 2022-12Raised $200M Series F led by Wellington Management at $3.7B valuation
  • 2013-02-14Founded in Paris by Florian Douetteau, Clement Stenac, Thomas Cabrol, Marc Batty
Revision history
21 July 2026
First assessed. Hard to displace. The text above is fixed as of this date and is not edited afterwards.
In the same band
Sources