French startup Dust has raised $16 million in a Series A funding round led by Sequoia Capital, allowing businesses to create custom AI assistants and share them with their employees to work more efficiently.
But what's interesting about Dust is how it differs from other companies working on enterprise agents, or AI assistants in general: Unlike consumer-facing tools like ChatGPT, Dust assistants are connected to your company's data and documents. For example, when you build a new assistant in Dust, it can connect to Notion pages, documents stored in Google Drive, Intercom conversations, and Slack.
At the same time, unlike most AI startups working on enterprise agents, Dust believes that companies should have not just one but multiple AI assistants, each performing a specific set of tasks and helping to solve common problems facing a particular team.
On a more practical level, support teams can use a Dust assistant that is aware of both the content of their knowledge base and past support interactions, allowing new team members on the support team to ask questions of their @supportExpert assistant and get the right answers.
Your HR team can create an AI assistant that can answer questions about company policies without having to search through complex Notion databases. You can also create another agent that can create job descriptions based on past job descriptions. Again, this empowers the entire company and frees up time for your HR team.
For engineering and data teams, the use cases are very simple: for example, Dust Assistant can understand your company's database schema and ask @SQLbuddy in plain English to create SQL queries about your customer base.
One last example: Sales teams can draft emails based on CRM data and a lead's general context, and the company offers an API if you need to build your own connector or integrate Dust Assistant with another tool.
Image credit: Dust
Dust is focused on building products that work for everyone, not reinventing the wheel. It's been a few years since ChatGPT launched, and by now most people are familiar with AI assistants (many even use them at work, even though it's against their company's policy). They know how to start a conversation, follow up with details, and ask the AI assistant to reframe their answer.
Using Dust isn't that different from how companies use the platform to build conversational assistants: Employees can access Dust's web interface or interact with the assistant directly in Slack, where they'll be @mentioned in conversations. Essentially, Dust wants to make generative AI a corporate communications tool that everyone uses every day.
The startup currently generates $1M in annual recurring revenue and has a growing portfolio of late-stage technology companies including Watershed, Alan, Qonto, Pennylane and PayFit.
Business banking startup Qonto estimates that 75% of its 1,600-person team uses Dust assistants monthly; at French health insurance unicorn Alan, 80% of the company uses its AI assistant weekly; and accounting tech unicorn Pennylane has built 86 custom assistants using Dust.
In addition to Sequoia Capital, some of the startup's existing investors are also returning, including XYZ, GG1, Connect Ventures, Seedcamp and Motier Ventures.
Taking a customer-centric approach also means that Dust does not create its own underlying models. When you build an assistant, you can choose which large language models to use for it. Dust is integrated with OpenAI (GPT), Anthropic (Claude), Mistral, and Google for Gemini models.
There are many startups working on enterprise platforms for building AI agents. Some names that come to mind are Brevian, Tektonic AI, Ema, Kore.ai, Glean, etc. Even Atlassian, the enterprise software giant behind Jira and Confluence, released their AI teammate Rovo. Let’s see if Dust has found the right go-to-market approach with a simple onboarding strategy.