Olivier Godement, drawing on experience with hundreds of enterprise deployments, says the top success pattern is pairing top-down leadership buy-in with a bottom-up tiger team of technical and institutional experts that starts small and scales. ✦ AI generated
Olivier Godement · BG2 Pod · 2025-09-11 · original ↗
starts at this moment · 17:30
“What does it take to build a successful enterprise deployment, a successful customer deployment and the counterfactual based on all your experience serving all these large enterprises?”
What I've seen being like clear leading indicator of success. Number one is like the interesting combination of like top down like buy in and like enabling like, you know, very clear group of like a tiger team, essentially, like, you know, the enterprise which sometimes a mix of like OpenAI, like, you know, enterprise employee.
verbatim transcript · starts at 17:30
17:30that point, I may have worked with like a couple of hundreds. I think. Couple of hundreds. So, okay, I'm going to pattern match. What I've seen being like clear leading indicator of success. Number one is like the interesting combination of like top down like buy in and like enabling like, you know, very clear group of like a tiger team, essentially, like, you know, the enterprise which sometimes a mix of like OpenAI, like, you know, enterprise employee. So, you know, typically,
17:57like, you know, you take like T-Mobile, like the top leadership was like extremely boring, like it's a priority. But then letting the team like, you know, organize and be like, okay, if you want to start small, start small, you know, and then you can scale it up, essentially. So that would be part number one. So top down buying and a bottom called a tiger team. Tiger team, you know,
18:15people like, you know, a mix of like technical skills and like people who just have like the organizational knowledge, like institutional knowledge, you know, it's really funny, like in the enterprise, like customer support, a good example, like what we found is that the vast majority of the knowledge is in people's heads. Right. Right. Which is probably like a thing that, you know, we have these like in general, but like, you know, you take a customer support,
18:39you would think that, you know, everything is like perfectly documented, etc. The reality is like the standard like operating procedures, like the SOPs are larger than people said. And so unless you have that tiger team, like mix of like technical and like, you know, subject matter expert, really hard like to get something out of the ground. That would be one. Two would be evals first. Like, whatever we define as good evals, like that gives like a clearly clear common goal
19:03for people to hit. Whenever, like, you know, the customer like fails to come up with good evals, it's a moving target. Essentially, you know, if you've made it or not. And you know, evals are much harder than what it looks to get done. And evals also oftentimes need to come up bottom up, right? Because all of these things are kind of in people's heads, in the actual operator's heads.
19:23Like it's actually very hard to have a top-down mandate of like, you got like, this is how the evals should look. A lot of it needs the bottoms up adoption. Right. Yeah. Yeah. And so we'd be dealing quite a bit of tooling on evals. We have like an evals product and you know, we're working on more to essentially solve like, you know, that problem or, you know, make it as easy as we can.