Agentic Engineer

Fusion Risk Management•,

6d•$150,000 - $165,000

About The Position

Fusion Engineering is building the agentic workflow that will power how our entire engineering organization delivers software. We have stood up a suite of AI agents capable of delivering features, enhancements, and bug fixes—and we are now looking for Harness Engineers to operate that system at scale and help it continuously improve. This is not a role about building the harness from scratch. It is about becoming an expert operator of it: steering agents toward quality outcomes, translating product requirements into precise agent-ready specifications, reviewing and validating AI-generated solutions, and closing the feedback loop so the system gets smarter over time. You will be the manager in the loop—the engineer whose judgment, communication, and technical depth determines whether a feature ships with confidence. Harness Engineers at Fusion work across the full delivery lifecycle: collaborating with product and design to scope requirements, crafting the context and prompts that direct agents, reviewing output for quality and correctness, and partnering with other engineering teams as embedded technical contributors. No two days look exactly the same—and the discipline itself is evolving fast. If you are energized by operating at the frontier of how software gets built, this role is for you.

Requirements

Deep, hands-on familiarity with AI coding agents and agentic workflows—including how to direct them effectively, where they fail, and how to recover when they do
Strong ability to write precise, unambiguous specifications: requirements, acceptance criteria, and constraints that agents can act on without additional clarification
Sharp technical judgment to evaluate AI-generated code and solutions—identifying gaps, regressions, and quality issues without necessarily writing the fix yourself
Experience with prompt and context engineering as a practical discipline: iterating on inputs to produce more reliable, higher-quality agent outputs
Ability to collaborate effectively across functions—translating between the language of product, design, and engineering stakeholders with ease
Comfort operating in ambiguity and adapting quickly as tooling, workflows, and best practices evolve in real time
Strong communication and documentation habits: the ability to articulate decisions, capture context, and make agent interactions auditable and reproducible
Bachelor’s degree in Computer Science, Engineering, or a related field (or equivalent experience)
5+ years of software engineering experience, with a strong emphasis on directing AI coding agents to deliver production-ready software—hands-on agent management is the core of this role, not traditional coding
Demonstrated track record of delivering features or enhancements using agentic or AI-assisted workflows—not just experimenting, but shipping
Deep understanding of how large language models and coding agents behave: context windows, non-determinism, failure modes, and how harness constraints improve reliability
Experience working in enterprise B2B SaaS environments with cross-functional product and engineering teams
Familiarity with CI/CD pipelines, version control workflows, and modern software delivery practices—even if agents are doing the implementation

Nice To Haves

Experience with Microsoft Azure or Salesforce platforms is a plus, but not required

Responsibilities

Translate feature requirements, enhancements, and product briefs into well-scoped, agent-ready specifications with clear acceptance criteria, constraints, and success conditions—including building Proof of Concepts (POCs) to validate approaches with Product Managers before full delivery begins
Operate a suite of AI agents to deliver working software—directing agents through the full delivery cycle from specification through review and validation
Review and validate agent-generated output for quality, correctness, edge case coverage, and alignment with product intent before solutions reach clients
Craft and iterate on prompts and context inputs that reliably direct agents toward high-quality outcomes across a variety of feature types
Collaborate with product managers, designers, and business stakeholders to refine requirements and ensure agents have the clarity they need to succeed
Work as an embedded technical contributor across product engineering teams, bringing agentic delivery capabilities to initiatives of varying scope and complexity
Identify recurring failure patterns in agent output and surface structured feedback to the platform team to improve harness guardrails and tooling over time
Champion quality and engineering standards in an AI-driven delivery model—advocating for rigor, transparency, and client-centered outcomes