Sources: Snowflake is in talks to acquire Reka AI, which builds LLMs for businesses, for over $1B; report: Reka was valued at ~$300M in a 2023 funding round
- Software company's venture arm has invested in the AI startup — Reka was started in 2022 by researchers from Google, Meta
Context & Ripple Effects
Reka emerged with funding to build custom enterprise models, with Snowflake among its investors in the earlier Reka funding round. The reported talks would move that relationship from venture backing toward potential ownership.
The proposed price would place a large premium on Reka’s reported 2023 valuation, illustrating how enterprise-focused model teams became strategic assets for data-platform vendors. Subsequent coverage says the acquisition talks broke down, underscoring that the report concerned negotiations rather than a completed transaction.
First-order effects
- Snowflake and Reka would need to manage customer, employee, and investor uncertainty while a potential acquisition is evaluated; no transaction is confirmed.
- A deal at more than $1B, if completed, would give Snowflake direct control of an enterprise-LLM team in which its venture arm had already invested.
Second-order effects
- The reported valuation step-up raises the strategic benchmark for other enterprise-model startups and for platforms deciding whether to partner, invest, or acquire.
- Snowflake’s cloud-data rivals face added pressure to show how their own AI offerings connect proprietary business data with model capabilities, rather than merely exposing third-party models.
Third-order effects
- If data platforms increasingly buy or tightly fund model developers, enterprise AI may consolidate around vendors that control both customer data environments and model distribution.
- The later collapse of these talks suggests that high strategic value does not eliminate integration, pricing, and independence trade-offs; partnerships and minority investments may remain important alternatives to acquisitions.
The trend: Enterprise data platforms are seeking tighter control over AI model capabilities, using investment and potential M&A to turn data estates into AI product distribution channels.