Glean: How to Outpace Competitors in Enterprise AI (and Frustrate OpenAI in the Process)

Intro

The question of whether AI, especially generative AI, increases productivity has been coming up more often, particularly with the billions invested in AI technologies. Meanwhile, Glean offers a tool that directly addresses productivity concerns in enterprises. Not driven by AI hype or ChatGPT, but by inefficiencies in large companies with over 1,000 employees, Glean’s founders – engineers through and through – have been using AI since 2019 to solve the problem of too many information silos and actually enhance productivity.

Not only did they raise a massive $260 million funding round in September 2024, reaching a $4.6 billion valuation – doubling what it was worth just six months ago – but they also made it onto several notable lists this year: IA40 by Madrona, Cloud 100 by Forbes, and “Don’t Invest in These 5 Guys” by OpenAI.

Why the distinction of being personally ousted by the biggest unicorn in GenAI (OpenAI, in this case)? Let’s explore in this new episode of the GenAI Unicorn Series.

In today’s episode:

• How it all started

• Why is it so hard to build an internal search engine? 

• What did Glean build? 

• But there was a problem: no one wanted to buy it

• The role of CEO

• Evangelize and ride the pandemic

• Impact of ChatGPT launch

• Tech Specs

• Hallucinations

• Financial situation

• Targeting and monetizing

• Competitors

• Conclusion

• Bonus: Resources

How it all started

You know what they say about secret agents? Once an agent – always an agent. Apparently, search engineers are of the same sort.

Arvind Jain – CEO and co-founder of Glean – worked at Google for a long time. His self-identification is as a “search engineer.” Even after he left Google and built Rubrik – a cloud data management company that focuses on providing backup, recovery, and cybersecurity solutions for enterprises – if there was a problem with searching for stuff, his professional ears would prickle.

And there were. Rubrik was growing rapidly, crossing 1,000 employees, but productivity had first slowed and then stagnated. An engineer who was writing 300 lines of code a day before was now writing 50 lines of code. It was really low. Same for salespeople. Arvind set an internal survey to find out what the problem was, how they could do better, and what things were coming in the way of employees doing good work.

The largest complaint was: “Hey, I can’t find anything in this company. I don’t know where to go and look for information, but I need that. And I also don’t know who to go and ask for help when I need help,” and so this was a big problem.

“So when people are saying, we can’t find things, the first thing that comes to my mind is, “Hey, we should actually have a search product,” thought Jain.

Reminiscing on his years at Google, Jain realized that the problem with internal search wasn't just a Rubrik issue – it was universal. Even at Google – one of the most powerful search engines – there was a huge problem locating the right information internally. So, he talked to a few more companies to make sure he wasn't hallucinating the problem, and then involved the best people he knew from the research engineering army.

Tony Gentilcore and Piyush Prahladka were Google veterans, while T.R. Vishwanath had worked at Microsoft and Facebook.

They became Arvind Jain’s co-founders at Glean, which they started building in 2019. 

Do you know how long they were in stealth mode? Two and a half years!

Why is it so hard to build an internal search engine?