Senior / Staff Software Engineer (AI Agents)

<div class="content-intro"><h3 class="" data-start="443" data-end="463">About Actively AI</h3> <p><a href="http://www.actively.ai/">Actively AI</a> is defining a new category: Intelligence-Led Revenue.</p> <p>Revenue organizations have always been bottlenecked on human capacity. Reps triage which accounts get attention. Context disappears at every handoff. On any given day, the vast majority of accounts have exactly zero people thinking about them.</p> <p>Actively addresses this at the structural level. Our platform deploys Per-Account AgentsTM across our customers’ TAM, working 24/7 to research, identify opportunities, and advance next steps without being asked. Leading enterprises including Ramp, Ironclad, and Samsara are already making this shift.</p> <p>Our co-founders are former Stanford AI researchers, and the team comes from Harvard, CMU, Berkeley, Brex, Scale AI, and Google. We've raised $68M from TCV, First Harmonic, Bain Capital Ventures, First Round Capital, and more.</p></div><h3><strong>About the Role</strong></h3> <p>We’re looking for a Senior/Staff Backend Engineer to architect and build large scale systems that power Actively’s GTM AI agents. These agents continuously reason over real-time customer and external data, generate next-best actions, and execute them reliably across systems with humans in the loop.</p> <p>In this role, you’ll design the scalable, event-driven, and fault-tolerant infrastructure that makes this intelligence possible — integrating with LLMs, orchestrating complex workflows, and ensuring every action happens with high reliability. This is a foundational role, shaping how production-grade agents operate at enterprise scale.</p> <h3><strong>What You’ll Do</strong></h3> <ul> <li><strong>Design and build the agents that power Actively’s platform. </strong>Model complex customer and business workflows into clear, reliable agent logic that reasons over streaming data, maintaining state, decides what to do next, and acts consistently across systems — continuously and at massive scale.</li> <li><strong>Develop orchestration and workflow primitives</strong> for agent and human collaboration. Translate nuanced, long-running real-world scenarios — multi-step workflows, customer-specific rules, and human-in-the-loop actions — into well-structured agent behaviors that are predictable, maintainable, and easy to extend.</li> <li><strong>Ensure reliability through good modeling. </strong>Design agents with clear state transitions, strong data contracts, and explicit guarantees — embedding observability and fault tolerance into their design so they behave correctly over millions of events and long-lived sessions.</li> <li><strong>