On-site, Canggu, Bali, Indonesia · 4–6 months

What does it take to read a noisy electrical impedance tomography (EIT) signal off a living body in real time — and close the feedback loop on-device before the moment passes? This is an AI & Automation Engineering Intern building the on-device inference and feedback loop for Lilia, a closed-loop EIT feedback device for female sexual health. Lilia builds AI that reads a living body in real time and acts on it in real time.

The payoff is proof employers hire on: a thing built that nobody asked to build. Everyone has a degree. Almost nobody has proof.

What you build

- You build the on-device inference loop for Lilia's EIT reads: firmware that turns raw impedance readings into a usable feedback decision.
- You work on acquisition across skin-electrode interfaces: the test fixture and calibration runs that separate real impedance shifts from contact noise.
- You are on the hook for the closed-loop control path: guardrails, logging, and fail-safe responses that keep an unsafe instruction from reaching the output.
- A validation harness comes out of this: a reproducible test set another engineer can run without you in the room.

The bar

- A closed-loop device built outside coursework, with the failure mode under real noise named.
- Claude Code, Hermes, or Codex driven every day on a real build, with the AI's errors caught before the build shipped.
- A firmware or control loop debugged to the point where the failure was repeatable, not random.
- Kept a hard inference problem on the table until the output was useful.

Who this is for

- Work until the trace is clean.
- A board on the bench beats a slide deck.
- A misbehaving control loop can hold attention for an afternoon — and it is not done because the error disappeared once.
- The map is dead.
- Ask the two-year question: do you want to leave with a thing built,



or a story about being busy?

Where you do it — and why here Thirty years building companies — New York, London, Tokyo, the Valley — and then it stopped mattering where the work happened. So the program runs in office in Canggu, Bali. A modern company costs twelve people and a toolchain — so being wrong six times a year is part of how we build.

The campus is 4,000 sqm, and the people who run the companies are on it every day. The bench sits in the same room as the person who can unblock the firmware. The cohort is fixed, the calendar is real, and the correction loop is in person.

What this location buys is attention, not a view. The glamour comes later, after the build is sent.

What happens after The internship is step one; paid work is the destination, for the ones who stay. The destination role and its terms — visa, pay, scope — are agreed before you commit, never after. The path is written, not guaranteed.

Terms

Unpaid under Indonesian law. A room is available on campus — rent paid by you. Epic Solutions PT hosts the on-site program and assists with the C22B internship visa process; the visa fee is yours. Daily lunch on us. Starting October.

Questions about the fit or the timeline? Tell us — we'll talk you through how the program runs.

Requirements

- A degree in computer science, engineering, or equivalent — you have proven you can finish something.
- Claude Code, Hermes, or Codex driven every day, with the AI's errors caught before the build ships.
- A thing built that nobody asked to build — a repo, a board, a firmware patch, a functioning prototype.
- Four to six months on-site.

How to apply Apply via the LinkedIn application field. Submit your CV and the repository, plus a sentence on the hardest failure mode you hit.

A finished CV was never the point. What counts is a thing you've built and the drive to build the next one. That is who we are looking for.

📌 AI & Automation Engineering Intern — Canggu, Bali (Bengaluru)
🏢 EX Epic
📍 Bengaluru