Software Engineer, AI Agents

Asana•San Francisco, CA

62d•$171,000 - $258,000•Hybrid

About The Position

We're building AI Teammates: agents that work like actual users in Asana and integrated apps. They triage bugs, respond to requests, draft project briefs, conduct research, and handle knowledge work across your team's workflows. Unlike chatbots, Teammates are shared team resources that build memory and context across all executions. They get smarter as you and your colleagues work with them. You'll help build the Teammates platform and experience, currently in beta with Fortune 500 customers and launching to GA next year. This represents Asana's shift from tracking work to getting work done. You'll ship agentic workflows, evaluation systems, and product features using our custom TypeScript framework with frontier LLMs. We have close partnerships with OpenAI and Anthropic and move at the pace of the industry. The work adapts quickly as agent capabilities evolve. This role sits on the AI Teammates team, working closely with PM, Design, and Research to deliver high-impact features. This role is based in our San Francisco office with an office-centric hybrid schedule. The standard in-office days are Monday, Tuesday, and Thursday. Most Asanas have the option to work from home on Wednesdays. Working from home on Fridays depends on the type of work you do and the teams with which you partner. If you're interviewing for this role, your recruiter will share more about the in-office requirements. This work is horizontal. It touches every part of the Asana experience. You'll solve problems around performance and reliability at scale, multi-user collaboration with agents, and persistent memory systems. The industry is moving quickly, so your work will adapt with it.

Requirements

Strong software engineering fundamentals including 4+ years of hands-on engineering experience. You've built reliable production systems and know how to operate them.
Excitement about AI and agentic systems with capacity to learn quickly.
Customer-oriented product mindset. Comfortable running experiments, making tradeoffs, and collaborating across engineering, PM, Design, and Research.
Clear communication and effective collaboration.
Comfort with ambiguity and rapid iteration. This work is at the bleeding edge and adapts as the industry evolves.

Nice To Haves

Previous applied AI experience like RAG, search, or eval frameworks is a plus.
Experience with or eagerness to develop skills in: agent/tool orchestration, evaluation frameworks, and production observability.

Responsibilities

The Teammates platform: agents that collaborate with users and each other across Asana and integrated apps
Agentic workflows in production (function calling, multi-step planning, memory, orchestration)
Evaluation harnesses, guardrails, and observability systems to improve reliability, safety, latency, and cost
Integrations with connectors and grounding sources (e.g., calendar, docs, tickets, retrieval systems)
A/B tests and experiments to measure user impact with cross-functional partners
End-to-end instrumentation and debugging for agent systems at scale
Contribute to engineering excellence through code reviews, mentorship, and pragmatic technical design.