Leads the team this role joins
Heads the group this role sits in
In-house, owns the opening
From your own Gmail and LinkedIn.
About the role
This is a principal-level software engineering role building an internal AI agent, open to remote work in the United States. It suits engineers with at least seven years of full-time experience and strong TypeScript skills. The agent is meant to be the main way people across the company get work done, from answering questions and analyzing data to running multi-step workflows. The work spans agent architecture, the interfaces people use, evaluation and observability, and shared building blocks other teams can extend.
What you’ll do
- Design and build the internal agent so it answers questions, analyzes data and runs complex multi-system workflows.
- Build the agent layer that interprets intent, plans work, selects tools and skills, recovers from failures and completes long-running tasks.
- Ship interfaces in Slack, on the web and on other surfaces, pairing chat with interactive views where chat alone falls short.
- Create evaluation, observability and feedback systems that measure whether the agent's answers and actions are accurate, useful and trustworthy.
- Provide shared primitives so other teams can add capabilities without fragmenting the agent or the user experience.
- Work out what it means for a company to be run through an agent, and carry those lessons into infrastructure customers can use.
What they’re looking for
- At least seven years of full-time software engineering experience.
- Strong TypeScript skills.
- Comfort across the stack, from conversational and interactive UI to APIs, queues, databases, permissions and production infrastructure.
- Care for AI interaction quality, including context selection, tool use, memory, latency, generated interfaces, and how a system signals uncertainty or asks for help.
- A record of adapting quickly to unfamiliar technologies, domains or responsibilities.
Nice to have
- Experience using or building agent harnesses such as OpenClaw, Hermes or Pi.
What’s on offer
- Competitive compensation including equity, with a San Francisco base pay range of 208,000 to 312,000 USD per year.
- Pay outside San Francisco may be adjusted based on employee location.
- Possible eligibility for a company bonus or variable pay program, depending on the role.
- An inclusive healthcare package.
- Mentorship and support for events that build skills and networks.
- Flexible time off.
- Equipment provided, plus a work-from-home budget.
Skills
Questions about this role
How much does it pay?
The San Francisco base pay range is 208,000 to 312,000 USD per year, plus equity and possible bonus eligibility, and pay elsewhere may be adjusted.
How many years of experience do I need?
The posting asks for at least seven years of full-time software engineering experience.
Is agent harness experience required?
No, experience with agent harnesses such as OpenClaw, Hermes or Pi is a plus but not required.