Agent Reliability Engineer, GTM

LangChain; San Francisco (CA); On-site; USD 150,000 - 190,000; Full time

Benefits offered by this job

401(k) plan

Meals on in-office days

Medical, dental, and vision

Flexible vacation

Job summary

LangChain is hiring for a GTM Engineering role focused on health, cost, and performance of the GTM Agent. You will own the agent running on LangSmith, working with Python 3.11, FastAPI, LangGraph, DeepAgents, LangSmith, and SQL-based data.

You’ll partner across GTM for high-leverage improvements and enterprise storytelling. You’ll monitor, triage, run evals, track usage, and build business metrics to show ROI.

Qualifications

• Experience building and running Python-based production apps.
• Strong SQL skills and data warehouse familiarity.
• Hands-on with LLM deployments, evals, and caching.
• SRE mindset: monitor, alert, and optimize reliability.
• Clear technical writing and knowledge sharing skills.

Responsibilities

• Monitor production health across GTM graphs and models.
• Triage issues from Slack/tickets and fix or escalate.
• Run weekly evals, investigate failures, add regression tests.
• Track cost, latency, and ROI by model and use case.
• Publish learnings and build monitoring for LangSmith.

Skills

Python

SQL

LLM apps

SRE/Operations

Technical writing

Tools

LangGraph

LangSmith

BigQuery

dbt

Job description

About Us

At LangChain, our mission is to make intelligent agents ubiquitous. We build the foundation for agent engineering in the real world, helping developers move from prototypes to production-ready AI agents that teams can rely on. We began as widely adopted open-source tools and have grown to also offer a platform for building, evaluating, deploying, and operating agents at scale.

With $125M raised at Series B from IVP, Sequoia, Benchmark, CapitalG, and Sapphire Ventures, we’re at a stage where we’re continuing to develop new products, growth is accelerating, and all team members have meaningful impact on what we build and how we work together. LangChain is a place where your contributions can shape how this technology shows up in the real world.

Today, our platform includes LangSmith (Observability, Evaluation, Deployment, Fleet, and Sandboxes), our open source frameworks (LangChain, LangGraph, and Deep Agents), and the newly launched LangSmith Engine for autonomous agent improvement. We have 100M+ monthly open source downloads, 6,000+ active LangSmith customers, and 5 of the Fortune 10 use LangSmith in production (+ 35% of the Fortune 500 overall), including teams at Klarna, Clay, Coinbase, Workday, Lyft, Cloudflare, Harvey, Rippling, Vanta, LinkedIn, Monday.com, Nvidia, and Bridgewater.

About The Team:

GTM Engineering builds the AI agents, systems, and automation that power how our go-to-market teams work. We partner across Sales, Marketing, Customer Success, Support, and other GTM functions to identify high-leverage problems and build solutions that improve speed, quality, and scale. Our work spans four core areas:

• Identify — find high-leverage GTM workflows where AI can meaningfully improve how we operate

• Build — design, build, and deploy production AI agents and automated workflows across GTM

• Enable — drive adoption through thoughtful rollouts, playbooks, best practices, and ongoing enablement

• Evangelize — share what we build and learn externally through content, demos, talks, and open source examples

About The Role:

You'll own the health, cost, performance, and business impact of the GTM Agent, and build the feedback loops that keep it improving. Because we build the platform we run on, you'll also operate the agent on LangSmith the way we tell customers to, and turn that practice into the reference story enterprises keep asking us for. You’ll work across Python 3.11, FastAPI, LangGraph, DeepAgents, LangSmith, Supabase Postgres, BigQuery, Anthropic and OpenAI models, and Slack and Next.js surfaces.

What You'll Do:

• Monitor production health across every graph, catching errors, slow runs, expensive runs, and silent failures before reps report them

• Triage incoming issues from Slack, tickets, and rep reports, fixing small things directly and routing the rest to the right owner

• Run the weekly eval suite, investigate failures, and turn real production bugs into permanent regression tests

• Track cost and latency by model, graph, use case, and role, and recommend concrete changes to model choice, reasoning effort, and caching

• Track usage and adoption per rep and per feature, and own the weekly health report the team runs on

• Build the business metrics that show leadership what the agent is worth, from reply rates and meetings booked to hours reclaimed and ROI

• Build our own monitoring and alerting on LangSmith, and write the "how we run our own agent" story for customers

What You'll Bring:

• Strong production Python and SQL, comfortable working in traces, logs, and warehouse tables

• Real experience running LLM applications, including tracing, evals, and prompt and cache mechanics

• SRE or production operations instincts: percentiles, SLOs, and separating noise from real pattern

• Healthy skepticism about metrics; you check what a number actually counts before you publish it

• Clear writing, and interest in publishing what you learn

• High agency; you notice what's missing and take initiative to build it

Nice to Haves:

• LangGraph or LangSmith experience

• Experience building an eval suite from scratch

• BigQuery or dbt

• Prior DevRel-adjacent writing

• Empathy for sales and go-to-market users

Salary: $150,000 - $190,000

Compensation Philosophy:

We offer competitive compensation that includes base salary, variable compensation for relevant roles, meaningful equity, benefits, and perks. Actual compensation and offerings will vary based on role, level, and location. Team members in the EU, UK, and APAC receive locally competitive benefits aligned with regional norms and regulations.

Benefits

Benefits include medical, dental, and vision coverage, flexible vacation, a 401(k) plan, meals on in-office days in the US and more.