Aside logo

MTS - Agent Harness

Aside

San Francisco, CAFull-timeSeen 1w agoStill listed 2 days ago

Most applications go out cold — see where you stand first. No sign-up to start.

Watch jobs like this.

At a glance

Compensation
No compensation found
Location
San Francisco, CA
Schedule
Full-time
Work Authorization
Visa required • Visa sponsorship

Olive lists jobs from US employers, including remote roles you can work from the United States.

Job overview

Aside builds a personal work agent that runs in the browser, using users' existing logins and sessions to automate tasks across the web. The team seeks an engineer to improve the agent harness, optimize performance, and run real‑world experiments while collaborating directly with founders and users.

Skills & qualifications

RequiredNice to have

Skills

TypeScriptDistributed SystemsSoftware ArchitectureEvalsAI AgentsChrome DevTools ProtocolPlaywrightAccessibility Trees

Benefits

Relocation Assistance

Full job description

About Aside Aside is a personal work agent built as a browser. Since it works where your accounts and context already live, it can securely use your logins, passwords, and existing sessions to complete work on your behalf across the web. Its browser agent scores 99.0% on Mind2Web, 40% smarter than Claude Cowork with Opus 4.7. Aside runs local-first, with memory and data protected on the user’s device. More than 50k people use Aside today, and we’re growing 25% week over week. The context Aside is a personal work agent that does your work for you. We want to make the agent faster and more efficient while improving its ability to use a computer. These goals go hand in hand. We need an engineer who can dig into our agent harness, run experiments, and make it better. You will:

  • Polish our browser automation framework and memory system. You can make everything from the ground up, from the agent loop, compaction, memory to browser automation framework. Here’s an article how we did.
  • Build an agent that uses your computer. Help Aside navigate browsers and desktop apps, understand what’s on screen, and complete tasks across them.
  • Optimize our harness. Reduce token usage, work around bot detection, and experiment with heuristics that replace long prompt instructions with predictable behavior.
  • Run real-world experiments. Build benchmarks and evaluation environments to test changes, catch regressions, and understand where the agent still fails on real tasks.
  • Build with users. Talk to them, understand where the agent falls short in their work, and turn their feedback into shipped improvements.

What we’re looking for

  • Use your brain. No benchmaxxing or /autoresearch. You read the code and prompts yourself, make changes line by line, and understand why each change helps.
  • An obsession with optimization. You care about making the agent more capable while reducing latency and cost. You track down wasted tokens and unnecessary steps, and measure whether your changes help.
  • Systems rigor. You understand how Chrome DevTools Protocol, Playwright, and accessibility trees work, and know how to put them to use.
  • Platform thinking. You think in primitives, not just features. You build solutions that solve today’s product needs while supporting future use cases.
  • Proactivity. You create your own tasks. When something breaks or feels below the bar, you propose a solution and ship the improvement as if the product were your own.

How we work

  • We’re a small, talent-dense team. Everyone has broad ownership.
  • You'll work directly with the founders.
  • You’ll directly talk to users through our community and interviews.
  • Our office is based in San Francisco.
  • We can help on housing (we have co-living house) and visa (depends on the situation).

Similar jobs, posted recently

Open roles like this one, listed in the last 30 days.

You've read the whole posting — now see how you match it.