Artificial intelligence touches every aspect of how we interact with information (and much more) these days. Today, a startup that’s building a company based on a particular application of that — how to apply AI to knowledge management in the workplace — is announcing funding as it finds decent traction for its approach. Sana Labs – which provides an AI-powered platform to help people manage information at work and then use that data as a resource for e-learning within the organization – has closed a round of $34 million after seeing the ARR rise sevenfold in the last year.
Menlo Ventures, the US venture capital firm, is leading the round for Stockholm-based Sana, with EQT Ventures and 25 angels and founders/operators also taking part. This is a Series B that values Sana at $180 million after the money.
There are many knowledge management, enterprise learning, and enterprise search products on the market today, but what Sana thinks she has found a unique solution to is a platform that combines the three to work together: a knowledge management-meeting-enterprise-research-meeting -Online learning platform.
The core of Sana is an AI platform and engine that connects to all the different apps an organization uses in the workplace – Salesforce, email, Notion, Github, Slack, Trello, Asana and anything you may need to capture, source or store information and communicate with others.
All data in these apps is ingested and organized automatically by the Sana (IA Magic) platform and retained as the information in these apps changes or develops. Then users who want to access the information go to Sana and ask for it in normal “human” language, like you would in a search engine. But at the same time, the data is used as the basis for e-learning modules for onboarding, training or professional development – modules created/designed either by people in the organization or by Sana itself.
This was not Sana’s original concept, which started out by just creating the main machine learning engine to organize information. But Joel Hellermark, CEO and founder of Sana, said that at first the startup was getting requests for the front-end — the part where people could easily query information and use it to create training materials and learning – so they also build that part. Learning can take the form of quizzes and polls, interactive sessions, etc., and when interactive Q&As are generated around webinars, like a kind of very resourceful, no-waste stew, the results of all these too be fed into the knowledge base for future reference.
The mix of knowledge management with research and e-learning means the platform sees very different engagement metrics, Hellermark said. “Sana is in continuous use, which is very different from a typical e-learning platform,” he said. “We are seeing active weekly and daily use” among the tens of thousands of employees at the more than 100 companies already using Sana, he added.
The technology itself is built and customized by Sana, but the models, Hellermark said, come from OpenAI, which has a “deep partnership” with Sana, in Hellermark’s words.
“We’ve been using their models since day one continuously, since before launch,” he said. This includes GPT, which – via ChatGPT – has been making waves among tech and media folks on chatty platforms like Twitter. Sana’s approach speaks to the longer-term evolutionary potential of AI.
“We believe there will be underlying models like OpenAI with the ability to refine them for specific areas,” Hellermark added. “For us, the focus is on the user experience on top of that.”
Hellermark describes himself as a lifelong obsessed with not only the importance of education, but also the power of AI to make its mark in space. But education comes in many forms – youth content, continuing education, adult learning and professional development are just some of the slices of the pie.
He said Sana chose to focus on the fourth of them for two reasons. The first is because of its convenience – there really isn’t anything else like it on the market today, but it’s definitely something organizations could use, given the overabundance of useful information. contained in the braintrust of an organization that operates on an inverse variation: the more it accumulates, the more difficult it becomes to exploit it.
The second reason for the focus on enterprise is the scalability factor: whereas education in the more traditional sense could clearly use tools to ingest a lot of disparate and fragmented information and make it easily accessible and form the basis of personalized learning modules for the individual, the fragmentation between age groups and school districts, not to mention countries and their own specific curricula, makes them a more complicated target – perhaps even more so in this moment, given the emphasis we’re seeing from startups and their backers, to focus on projects with sound unit economics, identifiable (and active) customer bases, and technology that’s already working at those purposes.
“The education sector is my greatest passion because if you solve learning, you solve everything,” he said. “But from day one we wanted to be a big business and it’s hard to scale that into K-12 because you have to adapt to different countries. Having a business approach helps us scale and helps doctors in engineers, product managers, sales people and everyone else – we can serve them all in more than 20 countries.
Crucially, that’s not to say that it won’t be a longer-term target, or that the traditional education sector wouldn’t or couldn’t be a receptive customer for technology like this – from Sana or from another startup – longer term.
Another important detail to consider is how Sana handles the quality of the information she gets. How does he decide — can he decide? — if the data it provides are correct, and what if there are several “answers” that are not consistent with each other?
“That’s what knowledge management is,” Hellermark said in response to the question. “You can have models that are just research, but that doesn’t take into account the need to verify knowledge and create journeys.” He said a “verification framework” is built into the system, which allows people to limit sources and other inputs that can be used by Sana, customers can choose to designate verified and accurate information, and choose whether users can access unverified information and classify information.
That’s not an entirely satisfying answer, to be honest, especially since accuracy is one of the most persistent issues around AI: what do you do if it’s not quite right , or outright wrong, or are you just using bad data?
As with the rest of the rocketry that is AI, however, this hasn’t been an issue so far hindering Sana’s growth.
“Over the past 6+ years, I’ve reviewed almost every other SaaS learning management system, and the best part about Sana is that they build a true knowledge management solution from the ground up, given how knowledge is captured in today’s knowledge economy,” said JP Sanday, the Menlo partner who led the investment. “Businesses are now more distributed, they are being asked to do more with less and cannot keep pace with innovation and must empower all of their employees. Sana is the only platform I have ever seen that can realize this vision. »
He added that the approach of people both drawing from the database and creating content around it creates a specific “organizational knowledge graph” that is more democratized than what you typically get in organizations.
“When I show prospects the product and they see the content creation experience along with the AI capabilities that help both authors and learners, they immediately know they’re looking at something completely different. – they see how much more expandable it is and how much more engagement they get from users,” he said.