Midship
AI for SOX testing
Founding Software Engineer
About the role
Midship is hiring a Founding Applied AI Engineer (Software Engineer) to help build the core agentic engine of our product.
You will work directly with the founders and CTO to design and ship agentic AI systems that perform real audit work end to end. This is not a research role and not a model-tuning role. The focus is on building reliable, production-grade AI workflows that solve concrete problems for customers.
You will be working on how AI agents plan, execute, and verify complex, messy real-world tasks. Audit work is the perfect problem definition for AI agents to solve, so you will be defining new best practices for the future of software.
This is an early, high-ownership role. You will shape core product architecture, influence how our agents behave, and help define how internal audit software is built going forward. You should be excited about working close to users, iterating quickly, and shipping things that get used immediately.
We are an in-person team based in San Francisco. This is one of our first non-founder hires.
About the interview
Our interview process is designed to reflect how we actually work. We care about practical problem solving, product thinking, and the ability to build reliable systems in ambiguous environments. We aim to keep the process efficient, respectful of your time, and focused on real work rather than abstract puzzles.
Below is what to expect.
Phone Screen
The first step is a short phone call with a Founder.
This is an informal conversation to get to know each other. We will talk about your background, what you are looking for in your next role, and what we are building at Midship. This is also your opportunity to ask questions and decide whether this feels like the right kind of company and stage for you.
We may touch lightly on technical topics or past projects, but the goal here is alignment and mutual interest, not evaluation under pressure.
Remote Technical Screen (1 hour)
The next step is a one-hour remote technical interview with a Founder.
This is a practical, hands-on session focused on how you approach real problems. The exact format may vary, but you should expect to work through a small feature or system similar to what we build at Midship. We are interested in how you reason, structure code, handle edge cases, and communicate tradeoffs.
The interview is language-agnostic, though our production stack is Python, SQL, and React. You are welcome to use any tools you normally use, including AI coding assistants.
We are not looking for trick questions. We want to understand how you build.
Onsite Interview (Half Day)
The final stage is an onsite interview at our San Francisco office.
This is a half-day session focused on practical feature building and working together in person. You will collaborate with the founders on a realistic problem, similar in scope and ambiguity to what you would encounter on the job. The emphasis is on problem solving, product judgment, and how you think through building something end to end.
We will also spend time talking about Midship, our roadmap, and what it would be like to work together day to day. This is as much about mutual fit as it is about skills.
We’ll provide lunch and make sure there is plenty of time for questions.
Timing and Logistics
We try to move quickly. The full process typically takes 1–2 weeks, and we are happy to accommodate candidates with competing offers or tight timelines where possible.
If you have questions at any point in the process, we encourage you to ask. We want candidates to leave the process with a clear understanding of how we work and what joining Midship would mean.
About Midship
Midship is building the platform of choice for internal auditors.
Internal audit today is weighed down by repetitive, procedural work. Teams spend huge amounts of time gathering evidence, ticking boxes, and producing documentation, instead of focusing on risk and advising the business. It doesn't have to be that way anymore.
Midship gives auditors AI agents that can autonomously perform large parts of audit programs, from evidence collection through to testing and workpaper creation. Our goal is not to replace auditors, but to dramatically increase their leverage so they can focus on higher-value work.
Over time, we believe auditors should not need to worry about how an audit is executed at all. They should be able to describe the risks they care about and the controls required to mitigate them, and let software handle the rest.
We care deeply about building products that are simple and intuitive. If a tool needs a long explainer video to get started, we think it has already failed. We build software that meets audit teams where they are today and feels obvious to use.
We are founded by experienced operators, not first-time founders. We met working together at a pre-Series A, YC backed startup and have been founding engineers ourselves. We have built real products, worked with demanding customers, and understand what it means to build from zero.
We are a small, in-person team in San Francisco, hiring our first non-founder roles. If you want to work directly with the founders, move quickly, and help define the future of internal audit, we would love to talk.
