Senior Full Stack Engineer

Notch is the AI operating system built for enterprises running complex, regulated operations. We deploy purpose-built agents that resolve real workflows end to end, compliantly and at scale, across insurance and financial services. Our platform connects customer and employee interactions with back-office systems to automate complex workflows — all delivered with the highest standards for auditability, explainability, and governance.
Backed by top-tier VCs, Notch has raised $45 million to date, including a $30 million Series A in March 2026 led by Headline, with participation from Lightspeed, Jibe Ventures, Illuminate Financial, Munich Re, and Phoenix.
About the Role:
We're looking for really good engineers who want to build great software.
Our platform runs hundreds of thousands of AI-driven conversations and workflows for enterprise customers who can't tolerate a wrong answer. The hard problems here don't sit in one layer — they sit in how the layers fit together: an LLM pipeline that has to be deterministic exactly where it matters, a data model where a leak between customers is a severity-1, long-running workflows that have to survive restarts and retries, and one product that has to ship into a shared cloud, a dedicated single-tenant environment, and a customer's own data center without becoming three different products.
You'll own problems all the way through from the product decision, through the data model and the backend, to what the user actually sees. Small team, real ownership, decisions that stick.
What You'll Do: Own features end to end: product conversation schema API UI

how it behaves in production.
Design the systems behind our AI agents — pipelines, orchestration, retries, guardrails, evaluation — so behavior is predictable and explainable.
Build backend services and data models that stay correct and stay fast as volume grows — across shared, single-tenant, and on-prem deployments.
Build the product surfaces our customers and our delivery teams use to configure, monitor, and audit agents.
Own your code in production: CI, deploys, observability, alerting, and the debugging that follows.
Work on how we ship, build and release pipelines, environments, and the infrastructure that lets one codebase deploy into very different customer environments.
Make the call on where complexity is worth it and where things should stay boring.
Raise the bar around you: code review, design review, mentoring, and the standards that keep a codebase alive two years from now.

Requirements:

You're a Good Fit If:

You have 4+ years of engineering experience, 3+ of them building real products across both frontend and backend.
You're strong in at least one modern typed language and ecosystem: TypeScript, Python, Go, Java, C#. We write TypeScript, and we expect you to be productive in it fast, but we hire for how you think, not for which language you last used.
You're genuinely full-stack - you can debug a React render, a slow query, and a stuck queue in the same day.
You know relational databases properly: schema design, indexes, transactions, and what happens under concurrency.
You've worked with distributed systems and know where the race conditions hide.
You're product- and solution-oriented - you ask what the customer is trying to do before you ask what to build.
You can take a vague, complex problem and turn it into an architecture that Just Works.
You care about your teammates, and you're looking to both learn and teach.
You work AI-natively. You use coding agents seriously, and you build the things that make them useful - skills, agents, evals, internal tooling, context that makes the whole team faster. This is how we work, not a side project.
You believe a small, dedicated team beats a big one.
You've built LLMs in production - not a demo, but something with evals, guardrails, and regression testing behind it.
You've worked in a regulated or high-stakes environment where auditability wasn't optional.
You take pride in frontend architecture - accessible, responsive, fast components.
An early-stage startup is what you're actively looking for: hard work, processes that aren't fully defined, and a lot of freedom.

Our Stack
We treat the stack as a set of tools, not as the solution — the specifics change, the thinking doesn't. The essentials:

TypeScript everywhere, strict, across a single monorepo.
Backend: Node.js, PostgreSQL, Redis, and a workflow-orchestration layer for everything long-running.
AI: the major model providers behind our own agent, retrieval, and evaluation layer.
Frontend: React, in real-time product surfaces our customers and delivery teams work in daily.
Infrastructure: multi-cloud, and deployed inside the customer's own environment where regulation requires it — so portability and data residency are design constraints, not afterthoughts.

Our Culture

Extreme ownership and accountability
High standards, fast pace, and lots of freedom
Minimal bureaucracy, evolving processes
People who want to grow beyond a narrow role definition
A small, commando-like team where your impact is immediate and visible
work from our office in the Bursa, Ramat Gan