Founding Platform Engineer: Finch
Finch helps companies grow their go-to-market by turning relationship-building into a repeatable system. We’re starting with small in-person events like sales dinners. Finch handles the planning, details, and follow-through so our customers can focus on the relationships, not the logistics.
We have built our waitlist of customers while working with our early design partners. Demand is very strong, and so are their results. We have some initial software we’re demoing that can do the job, and now we’re building the team that will take this through the exciting next stages as we put it in customers’ hands and iterate quickly.
Read This First
This is a founding engineering seat at a pre-seed funded startup. High everything: ownership, impact, intensity and a high bar. You'll be the first full-time engineer, and be close to customers as we keep shipping and not always behind-the-scenes. You’ll have the chance to work closely with the founder/CEO. This is a chance to shape how the product is built as well as make key technical decisions along the way. We have our initial funding and the backing of AI House (formerly known as the Allen Institute for AI’s “AI2 Incubator”).
Finch may feel different from places you’ve worked in the past. Together we’ll build a culture of trust, growth, psychological safety, excellence, customer-orientation, and inclusivity. The work matters, and our impact on the world matters, but who comes along for this ride and how we do that work also matter.
The Role
- Design and build the platform layer that lets non-technical builders and AI agents ship software collaboratively.
- Own infrastructure and DevOps for a fast-moving, AI-native engineering environment.
- Build and refine the internal tooling, automation, and guardrails that make AI-assisted development reliable rather than risky.
- Work directly with JP on product and technical direction. The technical direction is your job to make happen, but my experience is a resource to you.
- Ship constantly, learn the full lifecycle of Finch's product and customers, and take features from idea to production.
- Help define the engineering culture and practices for every engineer who joins after you.
Who You Are
- You live in modern AI-native development workflows. Fluent with AI agent orchestration for software building (software factories, and the like).
- You have hands-on experience with infrastructure and DevOps. You're going to be owning environments, deployments, and the systems other builders depend on.
- You're energized by ambiguity and comfortable working with autonomy once you understand the mission and have experienced the psychological safety afforded to you on this team.
- You’re ready to put in the hours if the mission, impact, and reward are in line.
- You care about building things that are safe and trustworthy enough for AI agents and non-technical teammates to build on top of, not just fast.
- You have a track record of shipping real, production software, ideally with some early-stage (Series A or earlier startup) or high-ownership experience.
- Location: in either the San Francisco Bay Area or Seattle, or in the process of relocating (sorry, at this stage we can’t help with relocation). And definitely will be spending some time in Seattle immediately after starting.
Nice To Haves
- Experience building platforms, internal tooling, or infrastructure that other builders (human or AI) work on top of including software factories, orchestration, or spec-driven (and validated) output.
- Early-stage startup experience, or a history of owning features end to end.
- Experience designing systems specifically to make AI-agent collaboration safer or more reliable (guardrails, evals, observability).
- Bonus: GTM, sales-tech, or events/hospitality domain familiarity.
What We Offer
- As the first full-time hire with a solo founder, you’ll find generous equity compensation in addition to the pre-seed salary (which will increase after a follow-on raise).
- Direct partnership with the founder and real influence over both the product and how it gets built.
- The chance to build an AI-native software factory from the ground up.
- Over time, investing in training towards your career growth.
How To Apply / Hiring Process
Our applications involves replicating the work to be done as closely as possible: doing a building exercise while using AI tools and discussing with the existing team.
How Finch Thinks About Technology, Including AI
I’ve been building software for decades. I know how we used to do things. Things are different now. AI-native is the future we get to live in now. If thinking about changing your title from “programmer” or “engineer” to a more accurate “builder” still itches, that’s OK, but that is how we are building our team. We’re not trying to review all the code that gets written, but we do need to ensure that it does what we want it to do and we don’t end up with regressions or breakages. Over time, product folks will use the building blocks that you make to move the product forward. You need to set them and their agents up for success. Ensure that security, data flow, and infrastructure are happening as they should.
We’re all still figuring some of this out. Capabilities change regularly, and we’ll keep being able to do more. We’ll bring on the latest approaches and technologies where we decide, together, that it makes sense.
We recognize that there are impacts to AI use. We are not blind to these and we care deeply. So ensuring that AI is applied in the right way, using local models where we can (they are getting more competent by the month), and making choices through that lens does matter to us.
Technical Stack
With the knowledge that things are still evolving and that we’ve done a lot of this “vibe-coded”, some thoughts:
- Currently it’s running on vercel, postgres database, with a vite+react frontend. I expect the database selection to keep, though we could add specialized things if needed.
- Long-term we have more experience with AWS and prefer that cloud.
- Through AI House we have lots of credits with AWS, GCP, Azure etc. (AI House compute + GPU credits are substantially more generous than any other place, and we also have a pattern of getting early access to new tools from the frontier labs.)
Many things can be changed to suit the needs that we have at the time.
How Finch Works Together / Our Values
We’re a small team, so how we work matters as much as what we build. Here’s what you can expect from day one:
- Problems, not tasks. I’ll hand you problems to the highest level your experience supports. Once you know where we’re headed, make the call and run with it. At my last company, one of my favorite feelings was going to start something and finding a teammate already had it rolling.
- Ship, learn, improve. We optimize for speed and learning, not perfection. Before over-engineering, ask: “Can we undo this later?” Mistakes are data. We talk openly about what didn’t work and keep stretching.
- Disagree with me. The people I trust most say, “I think you’re wrong.” Be direct, bring your reasoning, and expect me to ask why. When everyone in the room agrees, I get a little nervous. Before we debate, we make sure we can state the other side’s position fairly.
- Feedback is an optimistic act. I’m direct, and I tend to focus on what can be better. The spirit is “It’s not there yet, but we can make it better together. Let’s go.” Praise happens in public, and constructive feedback happens privately and live. I want your feedback on me too.
- Written and async by default. Status updates go in Slack, not meetings, and proposals get written down. Lead with your question, then the context. When we do meet, it’s short, has an agenda, and ends with clear next steps.
- Protected focus time. Builders should spend the vast bulk of their time on focused work. Block your calendar and turn off notifications.
- Show it early. Rough drafts and half-formed ideas are welcome. Early check-ins beat big reveals and save us both time.
- Commitments mean something. Commit to what you can deliver. If you’re overextended, tell me early so we can reprioritize together. The deadline is too late.
- Name your hunches. We use data where we have it and intuition where we don’t. We just label which is which.
- Unblock each other. Fifteen minutes of your time that saves a teammate half a day is always a good trade.
- Intense and sustainable. We’ll work hard, and boundaries matter. Ask for what you need; you don’t have to build a case for it, we’re all looking to bring our best.
- Trust and speed. Over time, we’re aiming for “no-look passes”: knowing where each other is headed well enough that a nod and half a sentence get the ball there.
A bit more about the founder: JP was previously co-founder & CTO at Brightflow AI (a UC Berkeley SkyDeck-accelerated startup) where we raised $20M through Series A.