Cohere launched an early access program for North, an AI platform designed to make daily work more efficient. The aim for Cohere North is to combine large language models, search and agents in one secure enterprise work and collaboration platform that can run anywhere.
With the move, Cohere, which was founded in 2019, is downplaying the LLM arms race yet could wind up being more effective for enterprise customers. The company refers to its North strategy as one focused on "meaningful AI. Other foundational model players are also working to diversify. Anthropic is adding collaboration features as it expands use cases for its Claude models. OpenAI is also expanding but has largely focused on expanding into search to compete with Google.
- Agentic AI: Three themes to watch for 2025
- Agentic AI without process optimization, orchestration will flop
The upshot here is that foundational model giants are going to have to surround their LLMs with applications. Cohere will have its models but is making a push to be an AI workspace platform to improve the speed and quality of everyday tasks. Cohere in July raised $500 million in funding for a $5.5 billion valuation and now has a more rounded offering.
According to Cohere, North will combine the following:
- Cohere models with other frontier models.
- Search discovery and automation.
- A multimodal AI search system called Compass to extract information from images, slides, spreadsheets and documents.
- Enterprise-grade security and privacy standards.
- A turnkey approach designed to speed up time to value relative to do-it-yourself approaches.
- The ability to be deployed in public and private clouds as well as on premises.
Cohere North, which is taking enterprise applications for early access, is a bet that AI returns are directly tied to workforce adoption. If teams have an easier way to use AI for everyday tasks, they're more likely to adopt it. Royal Bank of Canada and Cohere have partnered to co-develop North for Banking, a platform connects RBC's procedure and policy documents and processes with AI models.
In the blog post, Cohere said North is "optimized to run in private–including air-gapped–environments so that organizations can safely integrate all their sensitive data in one place." Cohere added that its vertically integrated stack will enable North to enable AI agents to perform complex tasks across enterprise silos.
Cohere is also taking aim at the retrieval augmented generation performance of much larger players such as Microsoft Copilot and Google Vertex AI.
What's next? Cohere appears to be going vertical. Cohere CEO Aiden Gomez said in a post:
"Since we developed each part of the technology stack underpinning North, the platform can be tailored to suit the unique needs of any business. This granular level of control is essential for customizing AI solutions to match each organization's needs such as industry-specific terminology and internal knowledge. Additionally, with our industry-leading focus on privacy and security, North is well suited for regulated industries where companies simply cannot risk their proprietary data."
In other words, the RBC partnership will likely be replicated across other industries.
