What you will do

• Own difficult problems from prototype through production, including the agent loop, backend, and infrastructure.
• Build agent-facing CLIs, connectors, idempotent API wrappers, and browser automation for enterprise platforms.
• Improve tool design, context management, guardrails, and evaluation loops.
• Own deployments across SaaS, self-hosted, on-prem, and air-gapped environments, including model routing, IAM, and upgrades.
• Work directly with live customer implementations where agent actions have real consequences.
• Use agents to build agents. We expect you to use them heavily instead of defaulting to manual work.

Who we're looking for

You are a strong generalist who has gone deep in backend engineering, infrastructure, ML systems, or another demanding technical area.

You already use agents heavily and have clear opinions about harnesses, evaluation, and context engineering. When documentation is incomplete, you would rather inspect the network traffic than guess.

Unfamiliar systems do not put you off. ABAP and X++ look like puzzles to solve. You think about injection, permissions, audit trails, and blast radius before something reaches production.

You have high agency, low ego, and can work through the ambiguity of early-stage 0→1 development.

Experience with SAP, Salesforce, NetSuite, D365, real-time voice systems, on-prem or air-gapped deployments, govtech, banking, defense, or founding and early-stage roles is a bonus.