All open rolesApply for This Role
AI & Agentic Developer
Design AI systems that expand expert capability while staying observable, testable, and accountable.
Why this role matters
You will explore where language models and agents can make demanding knowledge work better—and then turn the strongest ideas into dependable systems. This is a builder role for someone equally excited by a fast prototype and the hard work required to make it trustworthy.
What you’ll do
- Prototype and productionize AI-assisted and agentic workflows.
- Build retrieval, tool-use, orchestration, memory, and structured-output capabilities.
- Create evaluations for usefulness, accuracy, robustness, latency, and cost.
- Design human review points, safeguards, telemetry, and graceful failure paths.
- Integrate models with product workflows, enterprise applications, and structured data.
- Document system behaviour, limitations, architectural choices, and operational risks.
What you bring
- A strong software-engineering foundation and applied experience with modern language models.
- Fluency in Python and/or TypeScript.
- Hands-on knowledge of retrieval, tool calling, evaluation, and model observability.
- Practical awareness of privacy, prompt injection, model limitations, and responsible AI.
- Evidence that you can move from an experiment to something people can rely on.
What success looks like
- 01Promising AI ideas are tested quickly against real work and clear evaluation criteria.
- 02Production systems are useful, observable, and honest about their limitations.
- 03Trebor develops reusable applied-AI patterns instead of isolated demonstrations.
