MTS - Agent Harness Engineer

About Aside

Aside is a personal work agent built as a browser: it uses your existing logins, sessions, and context to complete work on your behalf across the web. Its browser agent scores 99.0% on Mind2Web, running local-first with memory and data protected on the user's device. More than 50,000 people use Aside today, with the product growing 25% week over week.

About the Role

Aside is hiring an Agent Harness Engineer to own the core infrastructure that makes its browser agent faster, smarter, and more reliable. This is a deeply technical individual-contributor role focused on improving Aside's memory system and browser automation framework, two of the hardest and most consequential engineering problems on the roadmap. You will work directly alongside the founders, run real-world experiments, and ship improvements that affect every one of Aside's users.

What You'll Own

Own and improve Aside's memory system, including the agent loop, context compaction, and memory architecture.

Build and refine the browser automation framework, including navigation, bot-detection countermeasures, and interaction reliability.

Reduce token usage and replace long prompt instructions with predictable, heuristic-driven behavior.

Design and run evaluation environments and benchmarks to catch regressions and measure agent performance on real-world tasks.

Talk directly to users, understand where the agent fails in their workflows, and ship improvements grounded in that feedback.

Apply prompt engineering judgment to make agent behavior more robust and efficient across diverse real-world conditions.

Must-Have

1+ years of real, full-time engineering experience; internship experience does not count toward this bar.

Hands-on experience building memory systems, browser automation frameworks, or closely related agent infrastructure.

Prompt engineering sensibility: the ability to read and rewrite prompts line by line, understand why each change helps, and replace verbose instructions with reliable behavior.

Based in San Francisco or willing to relocate; role is in-person five days per week.

Nice-to-Have

Prior founding engineer experience at a startup, having built and shipped in a zero-structure environment.

Active open-source contributions, particularly a GitHub project with 200 or more stars.

AI-native background with a track record of building on top of or improving LLM-based systems.