Evidence receipt / commitment
Published · transcript-backedDavid DeSanto: commitment
14 Apr 2024 Lenny's Podcast The GitLab way: Kindness, transparency, and short toes | David DeSanto (CPO)
“The next was we wanted to be both transparent, which fits into our conversation earlier and be focused on privacy with AI. And that means that we tell you what models we're using, we're source available so you can see how we're using them.”
Source trail
Everything needed to verify it.
- Speaker
- David DeSanto
- Attribution
- Verified speaker
- Claim type
- commitment
- Recorded
- 14 Apr 2024
- Publisher
- Lenny's Podcast
Transcript context
…Cool. Okay, just a couple more questions. I know you guys are doing some cool stuff with AI. Let's talk about that. What's going on with AI and GitLab? I would say we've taken a very unique approach to AI as part of software development. And the reason why we did that is that GitLab has a very unique position. We are a true DevSecOps platform. A lot of our competitors that people talk about are like a developer platform or developer experience type tool. And we know that if 25% of the SDLC is actually creating code, 75% is not, and we want to help all those people around the developer be it successful. And to do that, we set ourselves like three core tenets or principles for AI a couple of years ago. The first one was AI across the entire software development life cycle, we want to help product managers, QA teams, ops teams, security teams also benefit from AI. If you make your developer a hundred times more effective, it just means probably everything around them is going to break. And so to make them more effective, you've got to help everyone. The next was we wanted to be both transparent, which fits into our conversation earlier and be focused on privacy with AI. And that means that we tell you what models we're using, we're source available so you can see how we're using them. We tell you how they're trained. And the privacy part is we don't use your intellectual products for training and fine-tuning models. Our models are trained on data that is not yours, which means that you don't have more safety in using that. I think other companies have shown why that's important with leaks of customer data and we want to avoid that. I think the thing that's interesting, as a source available open core company, we're actually trusted by more than 50% of the Fortune 100. So if more than 50% of those top companies trust GitLab to secure their intellectual property, we had to do that with AI as well. And then the last one was AI efficiencies. We wanted to make it so there was a boost efficiency. GitLab ultimate, our top tier has a 7X boost on productivity. As from our customers doing a data analysis of their improvements, we want to take that to 10X as part of it. And so we think AI can give us that last little bit of a burst. Now I did mention a minute ago and me, I'll stop there to see if you have questions on those. And then I just want to talk a little about where we're going with AI because I think it's pretty exciting. I guess the only question is just for other companies, other product leaders thinking about integrating AI, working with AI, any tips or lessons you can share that might be helpful on their journey?…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.