Sr./Staff Software Engineer

Why Toma

At Toma, we build AI Coworkers for automotive that make it 10x more profitable for companies to sell cars.

We were the first AI receptionist for automotive dealerships. We've grown to several hundred rooftops nationwide, and now we're now taking those learnings up-market to Fortune 500 companies that sell cars and the financial products attached to them. AI has unlocked workflows that simply couldn't run at this speed or scale before, and we are capitalizing on the moment.

We're capitalizing because we've earned the right to. Our founders bring combined decades of experience across the automotive and AI industries: Monik (CEO) from Uber, and Anthony (CTO) from Lyft and Scale AI. More importantly, we've lived inside car dealerships for the better part of a year, doing oil changes alongside real service advisors and technicians, learning what actually breaks in a dealership's day. That ground-truth is the foundation everything we ship is built on.

The team we're building around that mission is small and deliberately so. We hire slowly and we hire for a specific shape of person: someone who takes real ownership, is hungry to chew on hard problems, is genuinely intellectual about the work, and is meaningfully optimistic about where this is all going. Our engineers and operators come from Rivian, Microsoft, Amazon, Applan, Bloomberg, and more, and every hire after this one will raise the bar again.

The Role

We need a Senior or Staff Engineer to own the technical direction across our user-facing surface area: the user-facing dashboard, the realtime voice runtime, offline jobs running at millions of tasks a day, and internal tooling. You'll write and review code every day, set the architectural direction, and help us scale to many times our current call volume without breaking what already works. This is a hands-on role.

What You'll Own

• Realtime voice infrastructure. Our voice stack was built entirely in-house using off-the-shelf speech-to-text, LLMs, and text-to-speech. You'll own end-to-end latency and reliability for live calls, where every 100ms is audible to the caller.

• Voice-to-CRM integration engine. Drive the technical roadmap for how Toma writes back into dealership DMS and CRM systems. Target 99.9% reliability for booking, customer records, and recall workflows.

• Agentic AI systems. Build the complex long-running online and offline agents that do real work on behalf of dealership personnel and corporate employees. Online agents that hold state across a conversation, take tool actions in the DMS, and hand off cleanly. Offline agents that run for hours or days, reasoning across thousands of records to surface deals, follow-ups, and exceptions a human would never catch.

• Production TypeScript across the stack. Next.js on the frontend, Bun-based services on the backend, PostgreSQL via Prisma, Temporal for durable workflows. Ship code that other engineers want to read.

• Engineering leverage. Mentor through code review, raise the bar in design docs, and partner with Product and Design to ship the right thing on a tight loop.

Your First 90 Days

• By day 30. Ship a meaningful feature in our enterprise web chat product, end-to-end. The chat widget is live on dealer and Fortune 500 sites and it's the fastest-growing surface we have. You'll find a sharp edge (latency, handoff quality, tool-use reliability, conversation grounding) and you'll close it.

• By day 60. Lead the design and rollout of an agentic workflow for a named enterprise customer. Concrete example: an offline agent that reasons across a Fortune 500's lead inventory and CRM history to surface and re-engage stale buyers, or an online agent that completes a multi-step financial-product flow inside chat without a human in the loop.

• By day 90. Be the architect-on-call for our enterprise chat and agentic surface. Code you've written or reviewed runs on every enterprise conversation, and you have a point of view, written down, on where this product goes over the next year.

What We're Looking For

• 6+ years building and scaling production full-stack TypeScript applications.

• Real experience with low-level Node.js or Bun. You have opinions about event loops, streaming, and backpressure.

• Comfort across the T3 stack (Next.js, React, Prisma, PostgreSQL, NextAuth, tRPC) or close equivalents.

• A track record of leading projects from ambiguous problem statement to shipped product.

If you don't tick every box but the work sounds like the work you want to be doing, apply anyway. Tell us what excites you about Toma; we'll figure out the rest.

Compensation and Benefits

• Competitive salary and real equity in an early-stage company.

• Free health, dental, and vision.

• Unlimited AI tokens (Claude, OpenAI, whatever helps you ship). We expect you to use them.

• Free daily lunch and dinner in our SF office.

• Weekly team outings and regular ride-alongs to dealerships so you see the customer.

• Unlimited PTO.

This role is in-person in San Francisco.