Research Engineer, AI Security - Build AI Security Agents

Tolmo; San Francisco (CA); Sur place; USD 180 000 - 250 000; Plein temps

Avantages offerts par ce poste

Competitive salary

Early-stage equity

Medical, dental, and vision

In-person in SF

Résumé du poste

Tolmo in San Francisco is seeking a Research Engineer on AI Security to help our agents plan, reason over a live graph of a customer’s production system, and decide what is worth testing. Your work sits between research and engineering and ships to production.

You’ll design evaluations, compare models and prompts, defend against prompt injection, and improve agent reliability at scale, collaborating directly with founders and security leads to ship impactful improvements.

Qualifications

• Hands-on experience building, evaluating, or improving LLM-based agents or systems in production.
• Rigorous, empirical approach; you design experiments and rely on evals over anecdotes.
• Strong production-quality coding in Python, TypeScript, or Go.
• Working understanding of security and how findings are validated.
• Familiarity with LLM security risks such as prompt injection, jailbreaks, and tool misuse.

Responsabilités

• Agent reasoning: improve how our agents plan, use tools, and reason over a live graph of code, cloud, and telemetry.
• Evaluation: build the evals, benchmarks, and graders that tell us whether an agent actually got better.
• Model and tuning: compare models, prompts, and agent designs on cost, latency, and quality.
• Model and agent security: defend our agents against prompt injection, data exfiltration, and manipulation.
• Reliability at scale: ensure long-running agents are dependable across thousands of runs.
• Learning from the field: study agent runs from customer engagements and turn failures into fixes.

Connaissances

LLM agents

Python

TypeScript

Go

Security basics

Experiment design

Evaluation metrics

Tooling

Description du poste

Tolmo in San Francisco is seeking a Research Engineer on AI Security to help our agents plan, reason over a live graph of a customer’s production system, and decide what is worth testing. Your work sits between research and engineering and ships to production.

You’ll design evaluations, compare models and prompts, defend against prompt injection, and improve agent reliability at scale, collaborating directly with founders and security leads to ship impactful improvements.