About OnRamp

OnRamp transforms B2B customer onboarding into a revenue driver. Our platform automates workflows, streamlines playbooks, and accelerates time-to-value, helping enterprises reduce onboarding time by up to 70%. Backed by leading investors (we just raised our $15M Series A) and trusted by Fortune 15 companies, we’re redefining how companies bring new customers online.

We’ve built the system of record for post-sale workflows, powering onboarding and expansion for enterprise customers including Fortune 15 companies. Now we’re evolving that foundation into something much bigger: an intelligent system that actually does the work.

We’re building an agentic platform that eliminates friction in the B2B customer relationship.

The Role

We’re hiring a software engineer who has figured out what most engineers are still catching up to: how to use AI to build enterprise-grade software dramatically faster — without sacrificing quality.

To be clear about what this is and isn’t: this is not an ML research role, and it’s not a prompt-engineering role. It’s a software engineering role for someone with real fundamentals who has made AI a core part of how they build — someone who directs coding agents, runs parallel workstreams, and ships in a week what used to take a team a quarter.

This is a high-impact individual contributor role on a small, elite team, reporting directly to our Head of Engineering. Our whole engineering organization is being built around AI leverage — you won’t be the odd one out fighting for tooling budget. You’ll be a founding example of how we work.

We’re deliberately not posting a checklist of frameworks and years of experience. We care about what you can build and how fast you learn, not what your last title was — we’ll calibrate level, scope, and compensation to you.

What You’ll Do

Ship enterprise-grade features end to end — from ambiguous product problem to production — at a pace that surprises people, with AI in the loop at every step

Direct coding agents (Claude Code, Cursor, Codex, or whatever wins next) as a force multiplier: parallel workstreams, asynchronous agents finishing work while you sleep, and your judgment steering all of it

Own the quality bar that makes speed sustainable: architecture, code review, testing, and guardrails that keep AI-generated code enterprise-ready — our customers include Fortune 15 companies

Work across the full stack of a modern web platform on AWS, including the agentic AI features inside our product

Help write the playbook for AI-augmented development at OnRamp — the workflows, tooling, and standards that make every engineer here faster

Partner tightly with Product and talk directly with customers, so what you build maps to real value

Who You Are

A software engineer first. You have strong fundamentals — system design, data modeling, APIs, debugging, testing — because AI amplifies judgment; it doesn’t replace it

You’ve fully integrated AI into how you build. Not autocomplete — agents doing real, delegated work. You can describe your workflow in detail and it’s clearly ahead of the curve

You have proof. You can point to production software you shipped with AI leverage — solo output that used to require a team — and explain exactly how you did it

You know where AI fails. You’ve developed the review discipline, testing habits, and skepticism to catch what agents get wrong before customers do

Enterprise-minded: you understand that reliability, security, and data integrity are features, and you build like people’s businesses depend on your code — because they will

High ownership and product-minded — you run toward ambiguity, make calls, and measure yourself on outcomes, not tickets

What Makes This Role Unique

The entire company is built around this way of working — AI leverage is the engineering strategy, not a side experiment

A lean team shipping at the impact of a much larger organization, on a real product with real enterprise customers

You’ll help define what AI-native engineering looks like — and build a career story very few engineers can tell yet