AI Acceleration Engineer

About the Role

Decagon is growing fast, and we're looking for an AI Acceleration Engineer to build the systems the company runs on and make AI a practical part of how every team works.

You'll extend the platform that already serves as our source of truth for contracts and customers, automate the billing, collections, and close work that still takes days of manual effort each month, and replace processes running on spreadsheets with systems people can trust. You'll also AI-pill the rest of the company, giving them the MCP servers, skills, agents, and artifacts to automate their own work.

This isn't a role about evaluating tools and writing decks. It's a hands-on technical role: you'll own what you build and keep owning it, working directly with the people whose processes you're changing. Because this space moves every few weeks, you'll be the one deciding what's worth adopting and getting it into people's hands. If you see manual processes as problems to engineer away, and have the range to do it, this role was built for you.

In this role, you will

• Own projects end to end, from requirements gathering and solution design through development, testing, deployment, and ongoing maintenance.
• Build and maintain the web applications and integrations internal teams run on, from existing systems such as our internal contract and revenue platform to new tools you design and stand up from scratch.
• Work with stakeholders across Finance, Accounting, Legal, People, and Recruiting to surface new use cases and areas to automate, and determine whether each calls for a custom build, an AI vendor, or a hybrid approach.
• Take on major workstreams as they come up: billing, revenue recognition, usage data ingestion, and reconciliation across our internal data, the ERP, and Salesforce.
• Build the interfaces that let others work with our systems through AI, including MCP servers, skills, and agent-facing tooling, and enable power users across the business to handle the simpler automations themselves.
• Measure and communicate impact back to leadership: time saved, cost reduced, errors eliminated.

Your background looks something like this

• You have 6+ years of hands-on experience building the internal tools, systems, and automations that business teams rely on — internal tooling, business systems, and/or revenue and finance operations.
• You have owned projects across the full lifecycle, from requirements gathering through deployment and ongoing maintenance.
• You're technical enough to build production-quality internal tools, but you're not looking for a product engineering role — the work here is solving business problems, and that's where you want to spend your time.
• You're a strong generalist who could go deep on billing one quarter and rebuild a recruiting workflow the next.
• You have sound instincts for how systems should be built — how data flows, where things break, and what it takes for someone else to maintain your work. Fundamentals matter more here than familiarity with any particular tool.
• You know when a workflow calls for an agentic/AI approach and when a deterministic pipeline is the better answer.
• You have hands-on experience with AI-native coding tools such as Claude Code, Codex, Cursor and have used them to ship production solutions. Coding experience is preferred: Python, TypeScript, or JavaScript; working with APIs; and deploying what you build.
• You build trust quickly with Finance, Accounting, Legal, and People, and can independently drive complex projects under ambiguity without needing formal authority to get things done.

Even better if you have

• You've led AI acceleration / transformation initiatives — you know the common failure modes and how to avoid them.
• You have a product mindset: you don't just build automations, you build things people actually adopt and love using.
• You have hands-on GCP experience — Cloud Run, BigQuery, Cloud Functions, Pub/Sub — and have built custom MCP servers, not just used them.
• You've worked on billing or revenue systems, or been through an ERP implementation