Internship · Engineering
AI Intern
01 About the role
You'd work directly with both co-founders — no layers, no ticket queue, no fetching coffee. We hand you real client and product work: building agents, wiring RAG pipelines, designing ComfyUI graphs, automating workflows in n8n, and deploying things people actually use. You'll see how an AI project travels from a vague client sentence to a running system, which is precisely the part college projects never cover.
02 What you'll work on
- Build and ship chatbots, LLM applications and RAG pipelines on real client data
- Design ComfyUI workflows and generative pipelines, custom nodes included
- Automate business processes end to end in n8n and in code
- Run evaluations, tune prompts, and chase down exactly why a model got it wrong
- Deploy what you build — Docker, Linux servers, APIs, logging and monitoring
- Sit in on client calls and see how scoping, pricing and delivery really work
03 Who this is for
- You're in your 3rd or 4th year of college, or equivalent
- You have AI projects to show — coursework, hackathons, side projects, open source
- Comfortable in Python, and comfortable reading docs you've never seen before
- Hands-on with at least one of: LLM APIs, LangChain / LlamaIndex, ComfyUI, n8n, vector databases
- You can genuinely commit 6–8 hours every day, consistently
- Self-directed — handed an ambiguous problem, you come back with good questions, then progress
04 What you get
- Direct mentorship from founders with 6+ years of AI industry experience
- Real production work in your portfolio — shipped systems, not sandbox demos
- Hands on the newest models and tools as they land, not a year later
- A written reference and recommendation on completion
- Flexible hours across time zones — we measure output, not a clock
- A clear conversion path, spelled out below rather than implied
05 The honest part
This internship is unpaid
We're an early-stage studio and we'd rather be straight with you than dress it up in equity language. There is no stipend during the first three months.
At the three-month mark we run a performance review. If your work has been strong and the studio's revenue supports it, we convert the role into a paid position. That second condition is real — we won't promise a salary we can't fund. For the right person, a pre-placement offer is genuinely on the table.
If either side decides it isn't working, you leave with the projects you built, a written reference, and no hard feelings. Please only apply if that trade sounds fair to you.
Tools, codebase, and your first change shipped to something real.
You take a feature or workflow end to end, with review rather than hand-holding.
Performance review. Conversion to paid, subject to your work and to revenue.
A pre-placement offer for the right person, ahead of your final year ending.