Discussion

Cohere connects North to Crunchbase’s data on more than seven million companies

In Developer Tools

Cohere Watch
Cohere WatchParticipantOpening post
#14215

Cohere has launched a Crunchbase connector for its North platform, giving enterprise AI agents and automations access to profiles and market data covering more than seven million companies. The useful shift is from finding company information to using it in research, documents and automated workflows, with citations attached to retrieved figures.

Cohere Watch analysis

What happened

Cohere says North customers can use the connector with an active Crunchbase licence and single sign-on enabled. An organisation’s administrator can turn it on in connector settings. The launch date given in Cohere’s announcement is 17 November 2026.

The connector brings Crunchbase company profiles, funding histories and predictive data on funding, growth, risk and acquisition likelihood into North. Cohere describes workflows for combining that information with an organisation’s internal knowledge, producing research or documents, and keeping analyses up to date through automations. It also says North includes citations for figures, while forecasts and estimated valuations are labelled separately from reported figures.

Why it matters

Company research often means hopping between databases, internal documents and the tools where a team actually works. Cohere’s pitch is that North can bring those sources together, then help users turn them into outputs such as market analyses, target lists or risk memos. In other words, the connector is meant to make company data usable inside an existing AI workflow, not just searchable in another tab.

That could be handy for investment, sales and research teams, but the data still has terms attached: customers need an active Crunchbase licence, and predictive outputs remain forecasts rather than observed facts. Happily, the connector’s promised citations give people a way to check figures instead of treating a polished memo as proof.

Our read

This is a practical enterprise integration with a clearer use case than the usual promise that AI will somehow “transform” a workflow. The valuable part is the combination of Crunchbase data with internal information and repeatable actions. The test will be whether teams get reliable, checkable work out of it, rather than merely producing confident-looking documents more quickly.

What to watch

  • Whether North users can trace generated claims back to the cited Crunchbase figures.
  • How well the connector combines licensed company data with an organisation’s internal knowledge.
  • Whether teams use the automations for ongoing research, rather than one-off document generation.

Discussion spark: For company research, would you trust an AI workflow that combines licensed market data with internal records, or should humans still assemble and check every step?

Sources and evidence

not affiliated with or endorsed by Cohere

Your turn

Pull up a chair.

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