Why Remote Breaks at Scale & the GTM Playbook Behind AI Agents w/ Darren Moffett, VP of Sales @ LangChain

This week Chris sits down with Darren Moffett, VP of Sales at LangChain, where he walked in 18 months ago with 5 reps and has scaled the team to roughly 60. It's his fourth open-source company after Cloudera, MuleSoft, and GitLab - two decades of selling technical products to technical buyers, from MuleSoft at ~1,200 people and doubling annually to GitLab months after its IPO. His company-picking screen for both: $100M+ revenue, growing north of 50%. It took meeting a dozen early-stage founders to break his own rule for LangChain.

Darren treats the conversation like an audit of his first 18 months in the seat: the pricing threshold reps started gaming, the comp plan he scrapped after living in constant exceptions, and the hiring profile he abandoned as the team scaled past 20. Plus what separates a technical seller from someone who can talk the talk, why remote work erodes at scale, and the "SDLC for agents" framework for selling a category that barely existed two years ago. It's the first-year retro founders usually only tell in private.

What pulled Darren to LangChain after three prior open-source runs: there's no competing with yourself. Most open-source companies monetize "enterprise-grade" features on top of the free tier, forcing customers to weigh how far they can get for free. At LangChain, open source covers the build; everything downstream, testing, evals, deployment, and production observability, is the paid LangSmith platform.

The nuance: open source drives "interest, not demand." It brings credibility, not conversion. The sale is getting customers to reorient to what happens after you build, and because LangSmith is neutral, they sell to teams building on ADK, Strands, or Vercel too.

Darren joined GitLab, the company that wrote the remote handbook, months after IPO and immediately noticed a lower level of trust and rapport between teams that he believes slows the business down. His verdict: "I wouldn't want to go back to a company that was 100% remote... the challenges outweigh the benefits."

But he splits it by function. At office-centric MuleSoft, his take was "if I've got salespeople in the office five days a week, that's a problem", sellers belong in the field. Where in-person matters most is early-career reps: 22-year-old Darren pounding phones from an apartment with four roommates loses focus in minutes, and the bullpen, overhearing calls blow up or land, is what accelerated his learning curve. "You curb their ability to learn and evolve" by forcing them remote.

MuleSoft's talent density came top-down from CEO Greg Schott. Asked how you protect culture while doubling headcount, his answer: "you just control who comes in the front door." Every candidate was scored on four criteria, cognitive ability, drive, track record of success, and good human, with each interviewer assigned one to dig into.

The differentiator was the final-stage exercise. Channeling Greg's goalkeeper analogy ("I want to see them on the field play"), candidates presented their current solution, not MuleSoft's, in a borderline combative session designed to test how they handle tension and reframe objections. At LangChain, non-traditional hires get an account-plan exercise instead, and some have even built demos of the product to prove aptitude.

His blunt read: customers want the best of both worlds, minimal commit, consumption pricing, but protection so consumption doesn't run away on them. LangChain sells access to the full platform with a minimum commitment, then draws down on actual usage, guiding customers to start conservative since agentic consumption is hard to predict.

The comp lesson is the transferable part. LangChain started on a net-revenue model and Darren scrapped it: accounts kept moving between reps (5 to ~20 in year one, 3x that now), someone always ate churn they never owned, and the team lived in constant exceptions. Now variable is split between new business and renewals on your own book, a carrot on both sides. Long term he expects pure consumption, "that'll probably panic some of my reps."

LangChain's internal AI leverage is a structure, not a tool. Everyone has access to Fleet, their agent builder, and the best internally-built agents bubble up. Then they embedded applied-AI engineers directly into the GTM org to scale the winners, including an agent that triages inbound, researches the account, and drafts outreach with a POV, all human-in-the-loop. Reps tweak the drafts and the agent learns to write in their voice.

His tool philosophy is deliberately hands-off: he spends zero time evaluating GTM tools. Adoption is bottoms-up groundswell, a rep brought Centralize to the table, 5-6 people validated it, and it rolled out to the whole team on the fly at kickoff.

(0:00) Darren's background + show open
(04:37) The MuleSoft talent machine: "you control who comes in the front door," and the four hiring criteria
(07:25) The goalkeeper test: making candidates demo their own product, borderline combative
(10:59) The bets on non-traditional profiles that didn't work, and the Deloitte one that did
(12:33) How to break into technical sales without the technical background
(16:47) Conscious competence: separating real skill from a lucky environment
(19:19) "Walk me through how you plan today": receipts over stories
(23:59) GitLab and why fully remote breaks at scale (and why early-career reps need the bullpen)
(28:19) His company-picking screen: $100M+ revenue, growing north of 50%
(32:57) Why LangChain: the open-source model that doesn't compete with itself
(34:12) LangChain 101: the SDLC for agents, and why you can't test them like traditional software
(39:42) Inside LangChain's internal GTM agents: inbound triage, account research, outreach in your voice
(44:43) FDEs and services, and whether heavy services means a weak product
(47:45) Usage-based billing: the commit-vs-consumption dance, and the comp plan he scrapped
(52:00) Year-one mistakes: the pricing minimums reps quietly gamed, and the hiring profile that had to change
(55:35) Rapid fire: the sales orgs he admires, AI tools, and the only real way to get his attention