Search is an obvious use case as companies look for ways to make the most of AI to improve employee productivity. After all, in most enterprises, data resides in multiple disconnected silos, and a combination of large-scale language models, search augmented generation (RAG), and enterprise-grade integration services make search possible. Not only can you answer more complex questions about all this data. It’s no wonder, then, that a significant number of startups and established companies (such as Atlassian’s Rovo) are vying for this market. One of the latest entrants is Doti AI, a Tel Aviv-based startup that today announced a $7 million seed funding round.
Doti AI brings together data from services like Confluence, Datadog, GitLab, Jira, Monday, Notion, Slack, Salesforce, and Zendesk and makes this data available on demand, for example through chat experiences in Slack. Through what Doti calls “autopilot.” AutoPilot actively monitors Slack channels and responds directly in that channel when it encounters questions that can be answered.
Doti co-founders Matan Cohen (CEO) and Opher Hofshi (CPO) told me that employees are moving away from seeing Doti in various Slack channels and starting to have one-on-one chats. It is also a way for Doti to quickly reach critical mass within the enterprise. Along with the service.
The two co-founders first met at Wix, where Cohen worked on the service’s infrastructure and Hofshi led a team of security architects. “When we left Wix[in 2023]we had no idea what we were going to do,” Hofshi told me. “Our first thought was to do one of our areas, what we’ve been doing for the past few years. For me it was cybersecurity, and for Matan it was infrastructure.”
During the ideation process, we realized that enterprise search remained an open problem. “The problems Doti solves are the same problems we felt at Wix,” Cohen says. And the solution, the founders argue, requires expertise in building integrations, but also a focus on security.
“What we believe is that every person, every group, every process, every company needs to have some kind of AI in their veins or processes. “It’s not a question of presence, it’s a question of when and how fast it will be deployed.” Companies will adapt,” Cohen said.
The team claims that its approach is quite flexible, even though other companies may appear to offer similar functionality at first glance. That’s because Doti, for example, was built from the ground up to allow businesses to bring their own large-scale language models to the service, while also helping them set their own security policies. “We know how to provide companies with very good solutions, and we’re as flexible as they want to be,” Cohen said. “Do you want to deploy your own LLM? Good luck. Do you want to deploy your own cloud? You can do this. Do you want to do AWS, do you want to do Azure, do you want to do GCP? Good luck, this is at the infrastructure level. It’s a big problem for the organization.”
The company’s $7 million seed round was led by F2 Venture Capital, with participation from a number of angel investors, including Ineffable Ventures general partner Jared Kasner and Sola Security CEO Guy Flechter.
TechCrunch has a newsletter focused on AI. Sign up here to get it delivered to your inbox every Wednesday.