Product Engineer

About the Company

They are a YC F25 company building the first agentic CRM - think Claude Code for managing your customers. They are not layering AI onto legacy software; they’re reinventing patterns for how work gets done when humans and agents collaborate.

They are backed by Y Combinator and built by a team of previous founders from AngelList, SAP, and IBM. They look for people who act like founders - driven by the mission, biased toward action, and committed to craft.

About the role

What you’ll do

You'll build the core of what makes it work and define how companies will run in an agent-first world.

Day to day, that looks like

• Inventing new patterns for how humans and agents work together
• Architecting systems for AI to run real business processes at scale
• Experimenting fast with new models, tools, and approaches as the space evolves
• Talking directly to customers, to translate their problems into features

The problems are genuinely unsolved: how should agents operate in the workplace? How do you make them reliable, trustworthy, useful? How do they interface with humans in a way that feels natural? What does agent to agent collaboration look like across businesses? They are running 100k+ agentic tasks per month and figuring this out every day.

Who this role is for

This role is for engineers who want to set the engineering foundation, not just inherit one.

You're independent. You don't wait around for specs or permission. You see what needs to happen and you make it happen.

You care about craft. The details matter to you across the whole stack, from data models to how a button feels when you click it. You ship work you're actually proud of and have a track record of doing so.

You have a high bar. You hold yourself to a standard higher than what's asked. You're comfortable making calls with incomplete information, and you'd rather ask forgiveness than wait for buy-in.

You're AI-forward. Not just curious - already building with it. You're a power user of Cursor or Claude Code, constantly exploring new tools, and you're genuinely excited about how AI changes the way you work.

Technically, you've been deep in Next.js / TypeScript for years. If you’ve built agents or multi-agent systems before, you’d be a great fit.

Founding experience, at any level, is a strong plus.

Why this role

This role gives you the rare chance to ship AI agents into production that are used by top startups to actually run their business.

They are building the system that businesses run on: a platform that defines how humans and agents collaborate. You'll be one of the first engineers shaping it.

This is a front-row seat to a high-growth startup at the bleeding edge of AI. You'll explore new models, tools, and approaches as they emerge - and actually ship them to production, not just prototypes.

You'll shape the engineering culture, the technical decisions, and the standards that define the company for years to come. The kind of experience that sets you up to become a technical leader or start something of your own.

Design is in their DNA. Their co-founder Andres is a second-time founder with a design background. Craft is a core advantage here, not a nice-to-have. You'll learn from people who've built companies before, not just worked at them.

They have real traction, funding from top VCs, and a team that moves fast without burning out.

Low ego, high trust.

Tech Stack

• Next.js / TypeScript
• PostgreSQL
• Vercel AI SDK
• AI-native tooling: they build with Cursor and Claude Code daily

Technology

Stack: Next.js, TypeScript, AI-native tooling (they build with Cursor and Claude Code daily)

What makes it interesting

They are not adding AI to a legacy CRM: they're building an Agentic System of Record from scratch. That means designing systems where AI agents run real business processes autonomously, trained by plain-English documents instead of complex configuration.

The technical challenges are genuinely novel: how do you build an Assistant that understands customer context across every tool? How do you make Agents reliable enough to handle thousands of workflows a week without hand-holding? How do you architect for a world where the AI does the work, not just suggests it?

Apply now!