Founding Software Engineer

Withq; San Francisco (CA); Hybrid; USD 180,000 - 240,000; Full time

Benefits offered by this job

Competitive salary

Equity

Hybrid in San Francisco

Direct access to founders

Job summary

Withq seeks a founding engineer to own core systems for its hospitality AI platform in San Francisco. You’ll be a founding engineer, architecting from scratch and partnering with the founders across product, GTM, and ops.

You’ll own the end-to-end stack including voice, AI agents, and integrations with PMS/POS systems. You will ship features rapidly, balance quality with speed, and help automate tasks currently done by humans until the team scales.

Qualifications

• 5+ years building and operating production backends.
• AI-native by default; coding agents and tooling daily multiplier.
• Innovating in agentic development to balance quality and speed.
• Strong design and architecture chops for scalable systems.
• Experience deploying LLMs in production (agents, tools, eval loops).
• Interest in telephony (SIP, carriers, PBX).
• Distributed-systems fluency: queues, idempotency, retries, state machines.
• High agency, low ego and fast iteration.

Responsibilities

• Own core backend systems from scratch and drive architecture decisions.
• Build and deploy AI-native agents and related tooling.
• Collaborate with product, GTM and ops founders; ship features end-to-end.
• Develop voice and telephony integrations with SIP and carriers.

Skills

TypeScript

AI-native development

Agentic development

System design

LLMs in production

Telephony

Distributed systems

Fast iteration

Tools

NestJS

Drizzle

BullMQ

AWS EKS

Aurora Postgres

React

Vercel AI SDK

Job description

We're looking for someone with a founder mentality — ready to own major product and business outcomes, not just tickets. This role is part founding engineer, part forward-deployed engineer, and part customer success when it's needed — until the CS agents we're building take that over. You'll wear all three hats, and help automate yourself out of the third.

We're building the AI operations layer for hospitality — an agentic workforce that meets guests on any channel (voice, SMS, web chat, email, WhatsApp), in English or Spanish. Our agents don't just answer; they have empathy and personality, and they sell — already closing bookings and upsells at rates above many human call-center reps. We're live in production with real hotels and ski resorts, with integration partners live and a new resort pilot in flight.

You'd be our second founding engineer — architecting core systems from scratch and working shoulder-to-shoulder with all the founders across product, GTM, and ops. Huge surface area, real ownership, design-doc to production.

• Voice AI — Sub-second voice agents (SIP, warm payment transfers). Latency is a feature.
• Agents that sell — Conversation design, booking + upsell flows, experiments that turn dialogue into revenue.
• AI agents & computer use — Agents that act, including computer-use agents driving legacy hospitality tools with no API.
• Hospitality integrations — Opera Cloud, DerbySoft, Hapi, and the rest of the PMS/POS/ticketing stack. Make systems you don't control behave.
• Synthetic NPS & call-quality analysis — AI grading every conversation into a weekly feedback loop.
• Automating the SDLC — Agent-authored code across the lifecycle, with the evals and guardrails that protect quality without giving up speed.
• High uptime, AI-native ops — Hotels run 24/7; build the kind of reliability they never notice.

What we’re looking for

• 5+ years — Building and operating production backends. Strong in TypeScript.
• AI-native by default — Coding agents and computer-use tooling are your daily multiplier. Core, not nice-to-have.
• Innovating in agentic development — Pushing the frontier of how AI builds software and automates the SDLC, balancing quality and speed rather than trading one for the other.
• Strong design & architecture chops — AI can write the code; it still lacks taste and foresight. You bring the system design and long-view it can’t.
• LLMs in production — Agents, tools, eval loops — not just demos.
• Telephony interest (or scar tissue) — SIP, carriers, PBXs, programmable voice.
• Distributed-systems fluency — Queues, idempotency, retries, state machines.
• High agency, low ego, fast — Ship and iterate over wait-for-perfect.

Our stack

TypeScript monorepo (NestJS, Drizzle, BullMQ) + React 19/TanStack — on AWS EKS, Aurora Postgres, Valkey, Istio, Auth0, Cloudflare. Multi-model (Anthropic, OpenAI, Gemini) via Azure Foundry/Vertex, with the Vercel AI SDK and Mastra powering our agents, and an evolving voice stack we keep at the very leading edge — Vapi, Deepgram, ElevenLabs, and others. We ship daily.

Bonus: voice/telephony depth · PMS/POS or travel tech · you love Gas City / beads (we lean on it heavily) · conversation/QA analytics · early-startup time · Spanish or French.

What we offer

Competitive salary, equity, and benefits. Hybrid in San Francisco (~3 days/week), direct access to the founders, and the most interesting problems in applied AI — voice latency, computer-use agents, agents that close revenue, and reliability at scale. And since our customers run everything from ski lifts to beach resorts, we hold the best deals close on a chairlift or a sun lounger, not a Zoom — pack your skis and a swimsuit.

withQ (Quickserve AI Inc.) is an equal opportunity employer.