Job description for AI Product Engineer Intern at Actionbook
Apply through Glints with your resume and one project you're proud of. Optional: a demo video, writeup, or a short note on an AI workflow you think should exist.
We're Actionbook — we let AI agents operate any website. People already use agents like Claude and ChatGPT to think; Actionbook gives them a fast, reliable way to act: research behind logins, fill forms, click through real sites, and handle repetitive tasks — all inside the user's own browser, with the user in control.
We're not building another agent. We're the tool your agent uses to get things done on the web — works with any MCP client.
We're an early-stage startup in Singapore. A small team, AI-first, ships fast.
About the Role
People connect their agents to Actionbook and then use it — they set things up, see what their agents are doing, and stay in control. Your job is to build that product: the interfaces, flows, and controls that turn our infrastructure into something people reach every day.
You'll work directly with the founding team and own features from idea to launch. This isn't a "small intern task" role, but you won't be alone either: real ownership, close support from the founders. We care about agency, taste, and shipping more than a perfect resume. You should love writing code and obsess over how it feels to use.
What You'll Work On
• Build user-facing product features end-to-end with TypeScript, React (Next), and APIs.
• Design the flows users actually live in, and make them clean and obvious.
• Prototype new ideas, then turn the best ones into real features.
You Might Be Fit If
• You've shipped things that people have actually used — side projects, hackathons, whatever.
• Solid full-stack with TypeScript and React.
• You work with AI coding tools, not around them — Claude Code, Codex, opencode, whatever gets you shipping faster.
• Taste. You notice when something's clunky and it bugs you.
• You can walk through what you built and what you'd do differently.
Nice to Have
• Strong design sense — Figma, design systems, or product prototyping.
• Experience with agent loops, tool-calling, or LLM-powered features.
• Familiarity with browser automation, Playwright, Chrome extensions, or computer-use agents.
• Open-source projects, public demos, or technical writing.
Details
• Location: Singapore, onsite.
• Commitment: At least 4 days/week, minimum 6 months.
• Compensation: Paid, with a path to full-time for strong performers.
• Start date: Rolling.
How to Apply
rough Glints with your resume and one project you're proud of. Optional: a demo video, writeup, or a short note on an AI workflow you think should exist.