What is the bull case for it being at the gateway layer?

I think when people talk about it at the gateway layer, I think there's maybe, like, two things that they're talking about.

So, one, it makes it way easier to port.

So you can take Claude Code.

People love Claude Code, right? If you tell people, "Hey, you can't use Claude Code, you have to use this other harness," there are other good harnesses out there, but let's pretend for a second that you love Claude Code, and you want to let them keep on using that, but you want to do some semblance of kind of, like, routing to maybe, like, a DeepSeek or, or an open source model or something like that, then the only way you can do that is at the gateway layer 'cause you don't control the harness.

Side note, this is another argument for open source harnesses that you can control and modify.

And then the second thing is when, when people talk about kind of, like, routing at the gateway layer, I think they're oftentimes implicitly assuming that there's a domain.

How do we actually think about that?

How did you turn this into a company, and how has LangChain the framework changed over time while LangChain the company also growing over time?

Yeah.

It's, it's a good question, and they've definitely both evolved a lot.

So maybe like going, going kind of like in a timeline view, you know, we launched LangChain October 2023-- uh, 2022.

We started the company, uh, in like January or February of 2023.

And from that point, we were working on our first kind of like commercial solution, which is LangSmith, so observability and evals, uh, for agents.

And, and, and the thing that we realized that actually has guided basically all of our, our products since then is that, like, the hard part of building agents is getting them to do what you want reliably.

I'm curious, how have you thought about the different domains of agents and then working backwards to the architecture they actually should have? I think different agents that do different tasks need to have different architectures.

So what is a deep agent, and then maybe we can get into talking about the enterprise, and why would an enterprise in particular care about that distinction?

Yeah.

So about a year ago we saw, uh, a, a few really interesting things.

So maybe even backing up, like, you know, three years ago, great.

Like, LLMs, you wanna connect data, you wanna connect these other things to them.

Fantastic.

How do you do that? Turns out it's really hard, and, and the best way to do that for, for different types of agents was actually pretty different.

... and how that plays into, and again, in the enterprise, you know, building that trust in what the agents are doing?

So can you walk us through that?

The core idea of deep agents came from looking at cloud code, from looking at deep research style things.

So OpenAI had a deep research.

Google had a deep research.

And from looking at Manus and basically realizing that they were all doing the same thing under the hood.

They were running in a loop and calling tools.

And, you know, we were around when when auto GPT came out and that was the same thing running the loop and calling tools that didn't really work.

And while you can do all of that with Landgraf or with deep agents, how should people in enterprise who are looking at sort of building something, how should they approach this?

Mm-hmm.

Then we have the run phase, and so this is really running agents at scale.

Um, we, we didn't do like deployment for LangChain in the early days because, to be honest, a lot of the things that people were building in, in, you know, March of 2023 were like really simple (laughs) compared to what they are now.

And so there just wasn't that much new infrastructure that was needed.

As we start going into like longer-running and more stateful agents, like that's really what our runtime is, is aimed at.

And then the, the manage aspect of it is something that we're just starting to do more of.

This is where like our insights agent comes in.

Mm-hmm.

What do they do? And how well does it work? And how long have you been

deploying your SDR agent? That's probably one of the newer ones.

Basically, what it does is we get a lot of inbound leads.

It does some research on who the people are, and it actually drafts an email to them if it thinks they're interesting.

So it uses the reasoning of the models to determine whether it's kind of like an interesting prospect for us.

It does some research on events that have happened to their company recently, and then it will draft an email.

And notably, for all of these positions, you're absolutely right that they're kind of like entry-level positions.

Just like back in the day, we wouldn't want somebody to spam, you know, and send 100 accidental emails could be embarrassing, could be annoying to our partners and customers.