Far away from Silicon Valley where much of the world’s AI attention is focused, Cohere has carved out a position for itself among AI leaders from Toronto, the tech capital of Canada.
The company differentiates based on its target market as well, focusing squarely on delivering AI to the enterprise while the company and its founding team describe lofty goals including “building the future of language AI” and “scaling intelligence to serve humanity.”
Company Background
Founded by ex-Google scientists, the company aims to build the most capable, advanced Large Language Model (LLM) applications, with a goal of providing exceptional usability and accuracy for enterprise use cases.
Today, Open AI’s Chat GPT is its primary competitor but Cohere is also in a race with Google, Meta, Microsoft, other established tech giants, and a plethora of startups.
As a privately held startup, the company doesn’t disclose revenue information, though at least one report placed its annualized run rate in the tens of millions of dollars.
The company said this week it has raised $500 million in new funding from Cisco, AMD, and Fujitsu, bringing its total raised to $970 million and raising its valulation to $5.5 billion. This more than doubles its previous valuation of $2.2 billion just a year earlier. The company opened a New York City office in March.
and CEO
Cohere has attracted other major investors including Nvidia, Salesforce Ventures, and Oracle, demonstrating strong confidence in the company’s potential to develop LLMs for enterprise use.
Two of its co-founders, Aidan Gomez and Nick Frosst, previously worked at Google and Zhang is also an AI expert.
Cohere ranks number 10 on Acceleration Economy’s Top 12 AI Ecosystem Pioneers.
Tech Innovations
Cohere aims to deliver higher performance and accuracy for enterprise applications, a different approach than some core competitors. Specific innovations and points of differentiation include:
- Enterprise Focus. Unlike OpenAI which has a consumer-oriented focus, Cohere has strategically positioned itself to serve the enterprise market. Cohere develops specialized software tailored to the unique needs and challenges of large businesses — such as automation, data analysis, and secure AI integration.
- Command Models: Advanced Language Models Optimized for Business. Cohere’s large language models (LLMs) are trained on business-oriented data like financial statements, reports, emails, and economic indicators, making them better suited for enterprise applications compared to more general, consumer-focused models.
- Embed Model: Multilingual Capabilities. Cohere’s models are trained using business terminology; they support over 100 languages, enabling an enterprise to use a consistent AI platform across global operations.
- Retrieval & Rerank Model: Retrieval-Augmented Generation (RAG). RAG is the process of optimizing the output of an LLM so it integrates data from a knowledge base outside of its training sources before generating a response. Cohere’s RAG extends the LLM’s capabilities to include specific domains or an organization’s internal knowledge base, all without the need to retrain the model. It improves LLM output so it remains relevant, accurate, and useful in various contexts.
- Flexible deployment options: Cohere provides the ability to bring its models to customer data, offering secure deployment choices — a crucial consideration for enterprises with sensitive or regulated data.
Strategic Ecosystem Partnerships
Strategic partnerships allow Cohere to continuously improve its AI capabilities, expand its reach, and provide its customers with the most advanced AI software. Its most noteworthy partners include:
Google Cloud
Cohere’s AI platform is built on Google Cloud, allowing it to leverage the scalability, security, and reliability of Google’s infrastructure.
Microsoft
Cohere has partnered with Microsoft to integrate its AI capabilities into Microsoft’s suite of productivity and collaboration tools. Cohere’s latest product, Command R+, was first available on Microsoft Azure and is now available on Oracle Cloud Infrastructure (OCI).
Oracle
Oracle is a product, cloud, and distribution partner of Cohere. Customers can use a native, GenAI service powered by specially trained Cohere foundational models and built on OCI. At the same time, Oracle is integrating Cohere models into its portfolio of business applications, including Oracle Fusion Cloud Applications and Oracle NetSuite.
Fujistu
Reinforcing its commitment to support a wide variety of languages and use cases, Cohere recently detailed an alliance with Fujitsu that calls for the two companies to develop and deliver enterprise LLMs and AI services with Japanese language capabilities. Fujitsu will be the exclusive provider of the services the partners develop jointly on the global market. They’ll offer secure private-cloud deployments to support customers in highly regulated industries.
Salesforce
Cohere is one of the first four companies to receive funding from Salesforce Ventures’ newly-launched $250 million GenAI fund. Today, Cohere’s technology can be embedded directly into Salesforce implementations, allowing quick access to user sessions and enhancing support workflows.
McKinsey
McKinsey, also a customer, collaborates with Cohere to build customized AI solutions that address specific business challenges around customer engagement and workflow automation.
Also, Cohere has partnered with leading AI research institutes like Mila to expand its research capabilities and improve accessibility of cutting-edge NLP technologies.
Customer Successes
Cohere is particularly strong in serving businesses that prioritize customer support, customer and employee onboarding, and personalized experiences. Its startup program also helps early-stage companies reach their full potential by leveraging AI to scale their business and gain a competitive edge at an affordable cost.
Some examples of customer success:
- LivePerson: The company’s Conversational Cloud platform is used by companies to engage with their customers through automated conversations at scale. LivePerson wanted to keep LLM conversations grounded and factual with outputs that match enterprise needs. The company worked with Cohere on re-ranking search results from its knowledge base then weaving the ranked results into a broader prompt for help in answering customer queries. Working with Cohere has helped LivePerson automate additional workflows, save money, and optimize resources by focusing staff on higher-value tasks.
- BlueDot: This customer provides unbiased real-time intelligence about global disease threats, which are disseminated through its technology platform. Prior to its work with Cohere, the information was only accessible via API calls tailored to data engineers and data scientists. Using Cohere Classify and Cohere Rerank, BlueDot made infectious disease intelligence available through natural language queries. This makes it easier for public health agencies, life science companies, and other stakeholders to monitor global threats, understand them faster, and take actions sooner.
- Borderless AI: This customer collaborates with Cohere to provide AI agents for human resources professionals to automate HR processes such as creating contacts, streamlining onboarding, and understanding HR laws in various regions. It uses Cohere’s RAG technology to serve the full range of HR documents across the 170 countries that Borderless AI serves. The company has stated that 50% of business users who sign up have been using AI tech to help them onboard Borderless AI faster.
Analyst Perspective
With a crowded field of competitors trying to find the next AI innovation, Cohere’s approach stands out. Rather than build a product for all potential AI users, Cohere has focused on the enterprise client. By creating products that allow businesses to augment the AI models with their own data, Cohere’s customers can achieve highly targeted insights and recommendations. And its multi-language capability means that a large enterprise can employ this product across the globe, regardless of a location’s native language.
When working as a Chief Procurement Officer, I often struggled with differentiating potential suppliers. There’d be a pool of potential partners but few could articulate clearly what set them apart from their competitors.
Cohere’s differentiator is clear. Buoyed by strategic partnerships and research collaboration, Cohere’s customer-centric innovation is enabling it to support and accelerate the development of new AI-powered software across a wide range of industries and applications. Clients don’t have to accept something that is “almost right” for their needs. By supplementing Cohere’s models with their own data, they can see the kind of unique results that their business require.
