The AI Customer Support Platform for Every Business.
Every SaaS founder gets the same advice on day one: raise money, hire fast, grab the market before someone with a bigger war chest does. In 2015, two engineers in Nantes decided to ignore all of it. They built a small live-chat widget, gave it away free for a year, then started charging a flat $25 a month. A decade later, Crisp is the software behind roughly 600,000 companies and about 250 million monthly conversations - and it has still never taken a euro of venture capital.
That combination - global scale, zero funding - is rare enough to be worth a second look. Crisp sits in one of the most crowded corners of software: customer support tooling, where Intercom, Zendesk, Freshdesk and HubSpot spend heavily to win seats. Crisp competes there with a team of around 25 people, most of them remote, and a product philosophy that reads more like a design studio's than a growth-stage startup's.
Strip away the branding and Crisp does something simple to describe and hard to build: it takes every channel a customer might use - the chat widget on a website, email, WhatsApp, Facebook Messenger, Instagram, phone - and pours them into a single shared inbox that a support team works out of together. Behind that inbox sits a built-in CRM that remembers who each customer is, a knowledge base for self-service, a ticketing system, status pages, campaigns, and developer SDKs for teams who want to embed it inside their own apps.
The newest and loudest piece is Hugo, Crisp's AI agent. Hugo reads a company's knowledge sources, works out what a customer is actually asking, answers it, and hands off to a human when it hits the edge of what it knows. Crisp says Hugo can autonomously resolve around 40% of incoming requests - the tedious, repetitive half of any support queue.
Crisp started where most support tools start: small businesses. The average Crisp inbox has around nine agents, which tells you the core customer is a growing company, not a call center. But over the years the roster crept upmarket. Names like Air France, Decathlon, Emma, Reedsy and Hoxton Mix now sit alongside the tens of thousands of SMBs, drawn by a rare pairing - enterprise-grade features at pricing that does not punish you for adding people.
Anyone who has run support knows the shape of the pain. Messages arrive from six different places. The same three questions get asked forty times a day. A customer emails, then follows up on WhatsApp, and now two agents are answering the same person without knowing it. Context lives in one tool, the conversation in another, and the customer just wants an answer.
Crisp's answer is consolidation plus deflection. Consolidation puts every channel in one place so nobody double-replies. Deflection - the knowledge base, the chatbot, and now Hugo - means a large share of questions never reach a human at all. Crisp has said that over half its customers never contact its own support team, which is less a boast than a design goal: the product is built so that the easy questions answer themselves.
The most unusual thing about Crisp is not a feature. It is the balance sheet. No venture capital means no board demanding hockey-stick growth, no pressure to bolt on lock-in, no incentive to price by the seat and squeeze. The founders have said out loud that they want to be "the Apple of customer support software" - a claim that would be easy to roll your eyes at if the design and the numbers did not partly back it up.
That independence shows up in ordinary decisions. Flat, per-workspace pricing instead of per-seat billing. Generous fair-use limits when the company moved upmarket, with existing customers grandfathered rather than repriced. A flat internal hierarchy where, as the company puts it, good ideas do not have to climb a ladder. It is a slower way to build, and that is the point.
Crisp did not arrive fully formed. It grew one module at a time, each one earning its place before the next appeared.
Hugo is the clearest signal of where Crisp is heading. Rather than sprinkling AI on top of the old product, the team rebuilt core parts of the platform to support an agent that can be trained, tested against simulated conversations, and switched live - all before it ever talks to a real customer.
Crisp sells subscriptions by the workspace: a free tier, then paid plans running from a small monthly fee up to a few hundred euros a month, each bundling a set number of agents. Extras - additional agents, more CRM contacts, message translation - carry modest add-on costs. Hugo is offered pay-as-you-go on top. There is no complicated per-seat math, which is precisely the pitch.
Because there is no outside capital, every hire and every server is paid for out of revenue. That constraint is also the strategy. A small, profitable, self-funded company can afford to think in decades, and Crisp's ten-year arc is the proof of concept.
In the market map, Crisp is the value-and-design player against heavyweights spending on sales and acquisition. It will not out-market Zendesk or out-raise Intercom, and it is not trying to. Its edge is a product that small teams can actually afford to grow into, wrapped in a company structure that lets it keep shipping without anyone pressuring it to sell, pivot, or cash out.
There is, worth noting, a completely unrelated American company also called Crisp that works in retail supply-chain data. This is not that one. The Crisp in this story is the customer-support platform from Nantes - the one built by Baptiste Jamin and Valerian Saliou, and still owned entirely by the people who built it.