Role description
ConduitAbout Conduit
Conduit is an AI agent platform built for hospitality. Hotels, vacation rentals, and resorts use our agents for guest communication and internal operations, and those agents run millions of workflows a day.
Hospitality is 10% of global GDP, and guests judge it by how fast their requests get resolved. Resolving one usually means working across several tools and several people. Our agents do that coordination.
Role: Agent Strategist
Some of our customers are property managers with no tech team. They hand us their guest support and operations, and we run them. Our agents do the work, and a remote team of virtual assistants picks up what the agents can't handle yet. We want the agents handling 95% of it.
You design, build, and measure those agents. You take a new customer's operation, turn it into agents and playbooks, show with evals that the agents work, and keep raising the share of work they handle. Every case a person has to pick up tells you what to fix in the agent next.
You also help decide what we build next. Once we run a customer's guest support and operations, we can take on much more for them: dynamic pricing and revenue management, getting their properties found in AI search answers, and accounting and owner management. You work out which of these to automate first and how agents and people will run it.
Why this role is different
You don't need to be an engineer, but you need to think like one. You break an operation into steps, pick the tools and context an agent needs, and test it before it talks to a guest. When something needs new code, you scope it with an AI coding agent and hand it to engineering.
You work from real conversations. Guest transcripts are messy, and finding where the agent went wrong is most of the job.
You own the results: automation rate, guest outcomes, and cost per customer.
You also get to work on some early robotics projects, from the applied side.
What the work looks like
Onboarding new customers: mapping how they run today, writing down their procedures, and standardizing them across customers where you can.
Building agents on our platform: their instructions, their tools, and the context they pull in.
Writing evals from real conversations, running them before every change, and watching production after.
Reviewing what the agents and the virtual assistants did, and deciding which work moves from people to agents next.
Scoping new products, from what to build to how we run it for customers.
Regular check-ins with each customer on what changed and what comes next.
Who you are
3+ years of experience. No hospitality background needed.
You have built, configured, or evaluated AI agents or automations that real customers used. Backgrounds in agent strategy, deployment, solutions, forward-deployed engineering, and support operations all fit.
You measure. You can tell us how you knew an agent was working, and what you did when it wasn't.
You like working with operators and the people doing the work as much as with the software.
This role is onsite in San Francisco. Candidates must be authorized to work in the US; we do not sponsor visas.
Compensation and Benefits
$140-220K base
0.10%-0.30% equity
Performance bonus
Comprehensive health insurance and 401k match
Dinner covered every day
Interview Process
Intro call with the founders
Agent case with the founders
Onsite in San Francisco, about 10 AM to 2 PM with lunch. We send you a few weeks of mock data from one customer 24 hours ahead. On the day, you walk us through what is failing in it, fix and test a live agent with one of our engineers, and pick which new product we should build first.
Culture fit conversation with the founders
References