SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 200,000 - 400,000 / annual
Decagon is a conversational AI platform enabling enterprises like Avis, Block, Chime, and Oura Health to deploy AI agents for customer support and operations across voice, chat, email, SMS, and other channels. The Research team develops the model and decision-making stack powering these agents, focusing on accuracy, robustness, and efficiency in real-world deployments.
As a Research Engineer focused on Safety, you will own end-to-end responsibility for making Decagon's AI agents safe, reliable, and controllable from evaluation through production. You'll identify real-world failure modes and build the models, evaluations, and safeguards that prevent them.
Key responsibilities include:
- Researching and building safeguards against prompt injection, unsafe tool use, sensitive-data disclosure, policy violations, and hallucinated commitments
- Building adversarial evaluations, simulations, red-team datasets, and regression suites informed by production failures
- Developing and deploying classifiers, judges, reward signals, post-training methods, and runtime safeguards for safer agent behavior
- Analyzing production traces and incidents to identify root causes, test mitigations, and measure impact
- Partnering with Security, Product, Infrastructure, Legal, and customer-facing teams to turn enterprise requirements into scalable safeguards and rollout practices
Required qualifications:
- 2+ years of AI/ML engineering, research, or AI safety experience
- Hands-on experience evaluating, post-training, or deploying language models or agentic systems
- Experience with modern post-training techniques (reinforcement learning, preference optimization, distillation, model routing, synthetic-data generation)
- Experience with adversarial testing, model red teaming, prompt injection, policy enforcement, privacy, or safe tool use
- Fluency in Python and modern ML tooling with strong experimental judgment and engineering depth to ship production systems
- Comfort owning ambiguous, high-stakes technical problems and making clear risk and product tradeoffs
Preferred qualifications include experience building safeguards for high-stakes or regulated enterprise workflows, and familiarity with human-in-the-loop review, incident response, or responsible rollout frameworks for ML systems.
About Decagon
AI / Data / Infrastructure; SaaS / Enterprise Software — AI agents for customer support and enterprise operations.
