Software Engineer

Looking for someone who aims to maximize their learning rate and is not afraid to solve hard problems. You should be self‑aware about your strengths and weaknesses, and able to adapt your skills to problems you haven’t seen before.

You are an AI power user, not a vibe coder: you routinely use tools like Copilot/Claude/Cursor to design systems, debug, and ship production‑quality code faster, not to avoid thinking. You can read and reason about unfamiliar codebases, design clean APIs, and own features end‑to‑end from spec to deployment.

You’re comfortable with hardware and low‑level details: you can build PCs from parts, diagnose hardware issues, and live in the terminal on Linux/macOS for debugging, automation, and devops‑ish tasks.

You care about efficiency: you write code that is fast and robust, think about performance and failure modes up front, and are willing to profile, benchmark, and refactor to keep systems reliable as they scale.

---

Interview Process

Interviewing for a software role at Synphony

Thank you for taking the time to learn more about Synphony’s interview process.

We’re building frontier physical AI systems for real robots in real environments. That means we’re looking for engineers who like owning problems end to end: talking to customers, working with real hardware, understanding messy real‑world constraints, and getting working systems into production. You should be comfortable around tools and machines, and already have a solid grasp of modern machine learning.

We care a lot about agency and proactive behavior. The people who thrive here are the ones who spot the right problems to solve, propose solutions without being asked, and can clearly explain their thinking and tradeoffs. Our process is designed to see that in action.

We also try to make the process as practical and fast as possible. You’ll talk directly with the people building and using our tools, see what we’re working on, and then show us how you’d contribute. Once you’ve come onsite, we commit to a clear decision within 7 days.

Below is what to expect.

First Call, What you’ve built

The first call is a lightweight conversation to understand mutual fit.

We’ll talk about your background, what you’ve built, and where you want to go next. We’ll share more about Synphony, the problems we’re solving in agricultural robotics and physical AI, and how our engineering team works day to day.

We’re especially interested in:

• Times you owned a project end to end (software, hardware, or both).
• Examples where you took initiative and drove a change without being told to.
• Your experience with machine learning in practice (not just coursework).
• Any hands‑on experience with tools, electronics, or robots.

This is also your chance to ask questions and decide if the kinds of problems we work on are compelling to you.

Second Call: Deep Dive into What We’re Building

The second call is more hands‑on and concrete.

We’ll walk you through what we’re currently building: the systems, data flows, robot stack, and product surface area you’d actually be working on. You’ll see real code, tools, and workflows—not just a high‑level pitch.

We use this time to:

• Share the technical and product context you’d be stepping into.
• Dig deeper into how you think about system design, reliability, and iteration speed.
• Explore how your ML background and practical skills could have the most impact.
• Hear how you’d approach and prioritize the kinds of issues we face.

You should come prepared with questions, ideas, and thoughts on tradeoffs. We want this conversation to feel like collaborating with a future teammate.

Onsite: Show Us What You Can Do

If we’re aligned after the two calls, we’ll bring you onsite.

The onsite is focused on seeing how you work in practice. Rather than abstract puzzles, you’ll work through practical problems that look like what we do every day. That may include:

• Extending or modifying an existing codebase that interfaces with robots or data streams.
• Debugging or improving a small system related to telemetry, control, or ML pipelines.
• Working through a scenario that involves both software reasoning and physical constraints.

You’ll have access to a normal development environment, and using AI coding tools is fine if that’s part of how you work. We care about:

• How you identify the right issues to focus on.
• How you explain your reasoning and tradeoffs.
• How you handle ambiguity and incomplete information.
• How you think about safety, robustness, and “what could go wrong” in the physical world.

You’ll also spend time with the team, in the shop and around the robots, so you can get a feel for how we collaborate, make decisions, and ship.

Timing and Decisions

We move quickly. From the time you start this process with us, our goal is to give you a clear yes/no decision within 7 days of the onsite.

If you’re working through other processes or have timing constraints, share that with us early. We’ll do our best to align, and if we can’t move fast enough, we’ll tell you directly so we don’t waste your time.

Our goal is to make the interview process feel like a realistic preview of working at Synphony: practical, fast, and focused on building real things with real robots in the real world.