Cohere connects PitchBook to North, giving agents private-market data

Cohere has released the PitchBook Data Premium Connector for North, its enterprise platform. The company's AI agents can now work with PitchBook's institutional data on companies, deals, funds and investors directly inside existing workflows. The connection model is simple: the customer brings its own PitchBook license and signs in, and the agents get access to the data. Every figure comes with a citation so a human can verify it, and access rights are inherited from PitchBook and North. • Available from October 26, 2026 for customers with an active license and SSO enabled • An administrator turns it on in the organization's connector settings • Use cases: M&A, strategy, finance and IR, sales, executive briefings

AI-processed from Cohere Blog; edited by Hamidun News
Cohere connects PitchBook to North, giving agents private-market data
Source: Cohere Blog. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceCohere Blog ↗
Editor-in-Chief
Key metric #1Available from October 26, 2026 for customers with an active license and SSO enabled
Key metric #2An administrator turns it on in the organization's connector settings
Use casesM&A, strategy, finance and IR, sales, executive briefings

Cohere has added a PitchBook Data Premium connector to North, its enterprise platform for AI agents. From October 26, 2026, companies that hold a PitchBook license and have single sign-on (SSO) configured can connect private-market data directly to agents and working documents. The scheme is called bring-your-own-license: Cohere does not sell the data, but connects what the customer has already paid PitchBook for.

For enterprise AI, this is a telling move. The main problem for agents in finance and strategy has long been sources, not the language model. An agent that writes a deal memo from the model's general knowledge is useless. An agent that pulls figures from PitchBook and matches them against the company's internal documents can already save an analyst several hours.

What data does the agent get

Cohere describes the available dataset as PitchBook's institutional data. It includes:

  • companies and their financial metrics;
  • deals and funding history;
  • funds;
  • investors and their activity;
  • market benchmarks.

A user asks a question, launches an analysis or requests a draft document. North combines the PitchBook data with the organization's internal knowledge base: strategy memos, presentations and past reports. This pairing of an external market picture with internal context is what makes the connector more useful than a plain database search.

Five use cases the connector is built for

Cohere highlights five areas where the connection should pay off right away:

  • Corporate development and M&A. Deal research and memo preparation, where PitchBook's financial data is matched against internal strategy documents.
  • Strategy and market research. Building and automating competitive landscape maps, tracking capital flows in a given niche, and estimating market size with cited data that can be refreshed at any time.
  • Finance and investor relations. Comparing company valuations, funding rounds and peer growth rates, so teams can prepare board updates and investor materials on current market data.
  • Sales and partnerships. Researching prospective customers and partners before first contact: recent rounds, investors, leadership changes, growth signals. North then drafts personalized messages from what it finds.
  • Executive briefings. An executive's question becomes a one-page, sourced brief within minutes, helped in part by North's document mode.

The list reflects the typical tasks of people who spend their days in PitchBook: corporate finance teams, strategy departments, IR teams and business development.

Access rights and number verification

Mistakes in financial materials are costly, so Cohere stresses control. The connector respects the permissions and access rights the organization has already set up in PitchBook and North. If an employee has no access to a particular dataset in PitchBook, the agent should not show them more. Every figure North takes from PitchBook comes with a citation. This lets an analyst quickly check the source instead of taking the model's text on faith. This human-in-the-loop approach is becoming the standard for enterprise agents in regulated industries.

How to enable it and where to start

The connection requirements are:

  • the company must have an active PitchBook license;
  • SSO must be enabled in North;
  • an administrator activates the connector in the organization's connector settings;
  • existing Cohere customers should contact their account team, and new North customers should contact sales;
  • those without a PitchBook license will first have to negotiate with PitchBook itself.

A practical tip for teams: it is best to start with one repeatable process where the figures need regular updates. Good candidates are a competitor map or a quarterly valuation benchmark for the board. Tasks like these show time savings right away and can be checked against citations. More complex scenarios such as M&A memos are better launched once analysts have gotten used to checking the links and have seen how the agent behaves on internal documents.

Story

Key Questions Answered

What data does the agent get?

Cohere describes the available dataset as PitchBook's institutional data. It includes: companies and their financial metrics; deals and funding history; funds; investors and their activity; market benchmarks. A user asks a question, launches an analysis or requests a draft document. North combines the PitchBook data with the organization's internal knowledge base: strategy memos, presentations and past reports. This pairing of an external market picture with internal context is what makes the connector more useful than a plain database search.

What are the key details on Five use cases the connector is built for?

Cohere highlights five areas where the connection should pay off right away: Corporate development and M&A. Deal research and memo preparation, where PitchBook's financial data is matched against internal strategy documents. Strategy and market research. Building and automating competitive landscape maps, tracking capital flows in a given niche, and estimating market size with cited data that can be refreshed at any time. Finance and investor relations. Comparing company valuations, funding rounds and peer growth rates, so teams can prepare board updates and investor materi

What are the key details on Access rights and number verification?

Mistakes in financial materials are costly, so Cohere stresses control. The connector respects the permissions and access rights the organization has already set up in PitchBook and North. If an employee has no access to a particular dataset in PitchBook, the agent should not show them more. Every figure North takes from PitchBook comes with a citation. This lets an analyst quickly check the source instead of taking the model's text on faith. This human-in-the-loop approach is becoming the standard for enterprise agents in regulated industries.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Autonomous AI Agents

Delegate Deep Research, Coding & Workflows to an AI Agent

Deploy autonomous multi-step execution with NomadNet Agent (agent.nomadnet.ai) — automated web research, data synthesis, code generation, and complex task orchestration.

Launch NomadNet Agent →NomadNet AI Platform
What do you think?
Loading comments…