We're hiring a Software Engineer Intern. We want someone who can take a design doc and a direction and ship it fast and bug-free, with little supervision. The best fit is someone genuinely into personal AI assistants.
The role:
- Ship features end to end - from a design doc and a direction, with little back-and-forth.
- Build the agent - tool calling, browser automation, email and message parsing, and memory.
- Own the backend - data models, services, integrations and deployment.
- Move fast - pragmatic architecture and tech debt are fine. Bugs, or a UX that falls short of what was asked for, are not.
What we're looking for:
The bar is high: we're looking for people who are a clear "hell yeah." A "maybe" is a no.
- Agency - give you a goal and you go. You unblock yourself and come back with something that works.
- Agent/Harness engineering - you've built the system around the model, not just called its API: tool calling, memory, context management, and keeping an agent reliable across long, multi-step tasks.
- Bias toward shipping - fast and bug-free, even if the code underneath is scrappy.
- Strong Bonus: A revealed interest in personal AI assistants - shown by your website, projects or work history, not just what you say.
Bonus: You've built your own assistant, or worked with OpenClaw or Hermes (or built your own version of one).
Example projects:
The kind of thing that would make your application stand out:
- A personal assistant on WhatsApp, Telegram or iMessage that reads your email and calendar and texts you first.
- A fork of OpenClaw or Hermes with skills you wrote, or your own agent built from scratch with tool calling and memory.
- A browser agent that does a real chore end to end, like finding products and filling a shopping cart.