Staff/Senior Software Engineer, Agent Engineering (Helios)
Remote • New York, New York • Full Time
Posted 1mo ago
Job Location
New York, New York
Tech Stack
Remote Work Policy
Fully remote
Employment Type
Full Time
Categories
AI Agent Engineer
About the job
We are seeking a Staff/Senior Software Engineer to join our Agent Engineering team, focusing on the Helios project. This role involves building and strengthening the foundations of our agent runtime, enabling the creation of new agent-driven products. You will develop systems that allow agents to manage long-running work from assignment to completion, ensuring agent activity is visible, steerable, and connected to relevant projects, people, evidence, deadlines, and deliverables. Additionally, you will construct the framework for agents to safely interact with internal systems and external services, and connect agents to our retrieval and knowledge systems to ground their reasoning, memory, citations, and generated artifacts in authoritative evidence. A key part of this role is improving the evaluation and release discipline for agent behavior, traces, and simulations to enhance quality while controlling latency, cost, and risk.
Responsibilities
- Own and strengthen the agent runtime for building new agent-driven products.
- Build systems for agents to manage long-running work from assignment to completion, ensuring visibility and connection to project details.
- Develop the tool and action framework for safe agent interaction with internal and external systems.
- Connect agents to retrieval and knowledge systems for grounded reasoning, memory, citations, and artifacts.
- Improve evaluation and release discipline for agent behavior, traces, and simulations to control quality, latency, cost, and risk.
Requirements
- Experience deploying agent systems in regulated environments (government, defense, legal, healthcare).
- Familiarity with policy and legal research requiring temporal accuracy, provenance, and citation precision.
- Experience operating self-hosted or multi-provider inference in restricted environments.
- Experience building agents that create durable artifacts or perform permissioned actions on behalf of a user.