AI Agent Developer
On-siteTaguig, Metro Manila, Philippines
Job Summary
Build spec-driven orchestrator and sub-agent architectures that manage sensitive data classifications and integrate tool gateways. Write full agent specifications defining purpose, capabilities, input/output contracts, guardrails, and tool boundaries with the rigor of safety-critical systems. Design and test system prompts and tool schemas to ensure accuracy across multiple tool choices, then prepare production evidence including test results and behavioral outcomes for review. Work with Product, Security, and Cloud Architecture to enforce designated guardrails and risk tiers, ensuring every agent operates strictly within approved boundaries.
Required Qualifications
- At least 4 years building production software
- 1–2 years specifically spent building LLM-powered agents
- Real hands-on time with at least one agent orchestration framework
- Working knowledge of the Model Context Protocol
- A real prompt engineering practice
- Comfort writing precise specs
- Enough AWS familiarity
Desired Qualifications
- Experience designing systems where one agent does something and a different agent checks it
- Any exposure to red-teaming or adversarial testing of AI systems
- Time spent in a regulated environment
- Some familiarity with how agent memory works
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.