Shawn Wang: belief

Transcript context

…Yeah. I mean, we have, you know, we don't really have telemetry on what people are using in terms of what parts of LangChain, the telemetry we have is like, you know, anecdotal stuff when people ask or have issues with things. A lot of it also is like, I think we definitely prioritize kind of like keeping up with the stuff that comes out. I think we added function calling, like the day it came out or the day after it came out, we added chat model support, like the day after it came out or something like that. That's probably, I think I'm really proud of how the team has kind of like kept up with that because this space is like exhausting sometimes. And so that's probably, that's a big focus of ours. The support, I think we've like, to be honest, we've had to get kind of creative with how we do that. Cause we have like, I think, I don't know how many open issues we have, but we have like 3000, somewhere between 2000 and 3000, like open GitHub issues. We've experimented with a lot of startups that are doing kind of like question answering over your docs and stuff like that. And so we've got them on the website and in the discord and there's a really good one, dosu on the GitHub that's like answering issues and stuff like that. And that's actually something we want to start leaning into more heavily as a company as well as kind of like building out an AI dev rel because we're 10 people now, 10, 11 people now. And like two months ago we were like six or something like that. Right. So like, and to have like 2,500 open issues or something like that, and like 300 or 400 PRs as well. Cause like one of the amazing things is that like, and you kind of alluded to this earlier, everyone's building in the space. There's so many different like touch points. LangChain is lucky enough to kind of like be a lot of the glue that connects it. And so we get to work with a lot of awesome companies, but that's also a lot of like work to keep up with as well. And so I don't really have an amazing answer, but I think like the, I think prioritize kind of like new things that, that come out. And then we've gotten creative with some of kind of like the support functions and, and luckily there's, you know, there's a lot of awesome people working on all those support coding, question answering things that we've been able to work with. I think there is your daily rhythm, which I've seen you, you work like a, like a beast man, like mad impressive. And then there's sometimes where you step back and do a little bit of high level, like 50,000 foot stuff. So we mentioned, we mentioned retrieval. You did a refactor in March and there's, there's other abstractions that you've sort of changed your mind on. When do you do that? When do you do like the, the step back from the day to day and go, where are we going and change the direction of the ship? It's a good question so far. It's probably been, you know, we see three or four or five things pop up that are enough to make us think about it. And then kind of like when it reaches that level, you know, we don't have like a monthly meeting where we sit down and do like a monthly plan or something.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.