Staff Software Engineer: Agent Runtime & Orchestration

Deliverance AI; United States; Remote; USD 170,000 - 250,000; Full time

Benefits offered by this job

Equity participation

Competitive cash compensation

Job summary

Deliverance AI is seeking a staff-level engineer to own the execution engine that runs agent workflows and multi-agent swarms in production. Based in London or remote, you will drive orchestration, scheduling, state management, and tool calls across the platform engineering group.

You will design runtime APIs, build human-in-the-loop gates, ensure reliability under partial failure, and mentor the growing platform team while shaping engineering standards and production readiness.

Qualifications

• Experience building reliable, production-grade distributed systems.
• Experience designing APIs with strong observability.
• Mentorship and cross-team collaboration.

Responsibilities

• Design and implement the runtime responsible for agent and swarm execution, scheduling, state transitions, and tool calls.
• Build human-in-the-loop gates and control points that allow workflows to pause, seek approval, or safely resume.
• Make orchestration reliable, observable, recoverable, and diagnosable under partial failure and changing workloads.
• Define stable runtime APIs and abstractions for applied-AI and forward-deployed engineering teams.
• Establish engineering standards for API design, failure handling, testing, operability, and production readiness.
• Provide technical leadership and mentorship as the platform team grows.

Skills

Distributed systems

Backend engineering

Orchestration services

Staff-level leadership

Job description

Deliverance AI is seeking a staff-level engineer to own the execution engine that runs agent workflows and multi-agent swarms in production. Based in London or remote, you will drive orchestration, scheduling, state management, and tool calls across the platform engineering group.

You will design runtime APIs, build human-in-the-loop gates, ensure reliability under partial failure, and mentor the growing platform team while shaping engineering standards and production readiness.