🔎 Spotlight #11: Inside Cohere
Canada’s homegrown AI powerhouse taking on the giants of the industry.
The founding story: bringing AI research to the real world
Cohere was founded in 2019 by Aidan Gomez, Nick Frosst, and Ivan Zhang in Toronto. If those names ring a bell, it’s because Gomez co-authored the famous Google Brain paper “Attention is All You Need” that introduced the Transformer architecture, the very backbone of modern AI. And no, this is not the Transformer architecture that you see in Michael Bay films with tons of explosions. Along with Frosst, who also worked at Google Brain, and Zhang, they set out to build an AI company focused not just on research, but on delivering usable products for enterprises.
Their vision was to build large language models that businesses could adopt easily, with a focus on privacy, security, and adaptability. Instead of chasing hype, Cohere leaned into practical deployment of generative AI in the enterprise world.
Wait — what is the Transformer architecture?
The Transformer is the innovation that changed AI forever. Before it, models struggled to understand context in long pieces of text. The Transformer introduced a mechanism called attention, which lets the model weigh relationships between words across a sentence or even a whole document.
Think of it like this: if you read the phrase “the ball she threw was red”, a Transformer can connect “ball” and “red” even though they’re separated by other words. This ability to capture context at scale is what made it possible to train today’s large language models, from ChatGPT to Claude to Cohere’s own Command R. Below is a chart that explains how a Transformer supports models.
What does Cohere actually do?
At its core, Cohere builds large language models (LLMs) and makes them accessible to enterprises in a secure, customizable way. Unlike consumer-facing chatbots, Cohere’s focus is embedding AI inside business workflows.
Here’s what that looks like:
Search and knowledge management: enabling companies to build semantic search engines that actually understand natural language queries.
Conversational AI: powering chatbots and assistants trained on an organization’s own data, giving employees and customers reliable answers.
Retrieval-Augmented Generation (RAG): Cohere’s Command R model specializes in this, combining an organization’s documents with LLMs to generate accurate, context-specific responses.
Privacy-first deployment: unlike some competitors, Cohere emphasizes enterprise needs, mainly data security, compliance, and adaptability. Ultimately, they are making it easier for Fortune 500s to actually adopt AI.
In other words, Cohere isn’t just chasing cool demos. They’re building the infrastructure for businesses to use AI in a way that is practical, safe, and scalable.
What’s new in Cohere’s world
Cohere has been making headlines as one of the most important Canadian AI companies:
Partnering with Oracle to embed AI into enterprise cloud systems
Working with McKinsey to provide generative AI tools for consulting and enterprise clients
Launching Cohere for AI, their research lab and non-profit initiative dedicated to advancing open AI science
Expanding their product suite with models like Command R, optimized for retrieval-augmented generation (RAG) use cases
Continuing to scale operations across North America and Europe
For me, it’s exciting to see a Canadian company not just competing in AI, but doing so on the global stage. Too often, Canadian startups sell early or fade quietly. Cohere is showing that we can produce world-class AI companies right here at home.
Funding and team at a glance
💰 Recently raised over $500M USD in funding at a $6.8B valuation
👥 Team of 800+ employees and growing rapidly
🌎 HQ in Toronto, with offices in San Francisco, London, and other hubs
🖥️ Website: cohere.com
The competition and how Cohere compares
Cohere’s biggest competition comes from the likes of OpenAI’s GPT-4, Anthropic’s Claude, and Google Gemini. Each of those companies has massive funding and brand recognition. But Cohere has a key edge: its focus on enterprise-first deployments and deep partnerships. By working with Oracle, McKinsey, and others, Cohere has created a distribution strategy that puts its models directly in front of businesses.
From my perspective, this is the right approach. Consumer AI products will come and go, but enterprises represent sticky, long-term contracts where reliability, compliance, and security matter more than flash. Cohere seems to understand that, and it gives them staying power.
My take
I find Cohere inspiring not just because of the tech, but because of what it represents for Canada. We’re often overlooked in global conversations about AI, yet here’s a Toronto-founded company competing directly with the biggest players in the world.
Spotlight #11 is a reminder that Canada is not just producing fintechs, SaaS plays, or niche startups. We’re producing companies that are shaping the future of AI itself. For anyone looking to work at the bleeding edge of tech, Cohere is an incredible place to watch.
🧭 Cohere Roles — Now Hiring
There’s a ton of more roles open at Cohere, but leaning more towards the technical side of tech rather than the business side. However I encourage you to keep in tune with future LaunchPad job posts to see any new exciting roles that pop up at Cohere and many other cool firms!
That’s a wrap on Spotlight #11
If you made it this far, thank you for reading! I hope these spotlights help you discover companies worth your time, and make the job hunt feel just a little bit less overwhelming. Keep an eye out for next Tuesday’s LaunchPad job drop, and if you know someone looking to break into tech or level up, feel free to share this with them too.
Here’s to building something cool, or better yet, joining something cool. 🚀
