Cohere makes enterprise play with AI workspace platform
The startup is evolving into an enterprise platform provider. Its success depends on strategy and its ability to clearly define what its platform does.
AI foundation model provider Cohere is looking to be an alternative to Microsoft and Google in the enterprise.
The vendor on Jan. 9 launched an early access program for North, a new AI workspace platform that combines large language models, search and agents to help enterprise employees integrate AI into daily workflows, according to Cohere.
North lets organizations customize and deploy AI agents that help users find relevant information, conduct research and analysis, and perform complex tasks. Any employee can use North to create, customize and share an AI agent, Cohere said. The AI workspace platform uses Cohere's multimodal AI search and discovery technology, Compass.
Cohere is positioning the platform as an alternative to Microsoft Copilot and Google's Vertex AI Agent Builder.
More than just a model provider
With North, Cohere is seeking to capitalize on the recent popularity of AI agents and helps Cohere move from being just an AI foundation model provider into an AI platform provider for the enterprise.
"Models are very important, but models are an underlying detail," said Arun Chandrasekaran, an analyst at Gartner. "At the end of the day, customers are looking to take these models and build custom workflows and custom applications, and Cohere is trying to offer them that opportunity through their North platform."
The move toward being a platform provider makes sense for Cohere, he said, especially as the vendor struggles to gain headway ahead of other model providers like OpenAI, Anthropic, Google and Meta.
Cohere is still not mentioned in enterprise inquiries, in the same league as the top model providers.
Arun ChandrasekaranAnalyst, Gartner
"Cohere is still not mentioned in enterprise inquiries, in the same league as the top model providers," Chandrasekaran said.
Therefore, the AI vendor has recognized the need to pivot toward building more specialized models, such as its model for enterprise search.
However, the need to pivot is also about making money and gaining ROI in a market in which many foundation model providers have struggled to generate profits.
"There's a lot of investments going in, and you have to have money coming out," said Mark Beccue, an analyst at Informa TechTarget's Enterprise Strategy Group. He added that the best strategy for model providers like Cohere, Anthropic and OpenAI is to remain independent from cloud giants such as Oracle, AWS and Microsoft. "If you stand alone, you're going to have to make money at some point. They haven't done that yet."
"How are you going to make money with just an LLM, I don't know," Beccue continued. "It doesn't look good to me."
A different strategy
For Cohere, the pivot will only be beneficial if it can differentiate itself from the giant cloud vendors.
"The success or failure of North in the marketplace and Cohere moving forward will depend upon their go-to-market strategy and how well they partner with others so that they can have those strategic conversations with the people who are making these buying decisions," said David Nicholson, an analyst at Futurum Group.
Many enterprises already have relationships with a hyperscale cloud provider.
"[Cohere] could have a superior offering but be completely dismissed by the marketplace in favor of Google Cloud Platform or Azure simply because of name recognition and the scope of relationships that companies have with those other giants," Nicholson said.
While Cohere's messaging about North meeting data security and privacy standards appears on par with what enterprises are looking for, the vendor's strategy is really what will determine whether it's successful, he added.
Cohere says North is designed to meet stringent security and privacy standards. Enterprises can use the platform without risking their data because it runs in a private environment.
Unclear definition
Another challenge for Cohere is making sure it is clear in explaining what North does, as well as its definition of AI agents, Beccue said.
While the vendor has specified finance and HR as prime vertical applications for North agents, it's unclear what function those agents will perform.
"You're skipping over a description of what an agent does," Beccue continued. "What does a Cohere North agent do at the higher level?"
Cohere says enterprises can use North to deploy AI agents to help them find relevant information and connect to workplace tools and applications that employees regularly use. North can be deployed in a private cloud or on-premises.
Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.
