Runtime safety filters that detect or block harmful, off-policy, or unsafe LLM outputs before they reach end users; the production complement to alignment research. Primary AI lifecycle stage: evaluation and serving infrastructure.
1,219 active AI roles across 172 companies in our index reference Guardrails as of today.
The companies with the most active Guardrails listings are: Capital One (96 roles), Amazon (84 roles), JPMorgan Chase (74 roles), Google (55 roles), OpenAI (49 roles).
Guardrails primarily belongs to the evaluation and serving infrastructure stages of the AI lifecycle. In current hiring, Guardrails roles concentrate at: agents (74%), serving infrastructure (8%).
The sectors with the most active Guardrails hiring are: Enterprise, Big Tech, Banking.
Runtime safety filters that detect or block harmful, off-policy, or unsafe LLM outputs before they reach end users; the production complement to alignment research.
Primary AI lifecycle stage: evaluation and serving infrastructure.
As of today, 1,219 active AI roles across 172 companies in our index reference Guardrails. Hiring concentrates at the agents (74%) and serving infrastructure (8%) stages. Most common sectors: Enterprise, Big Tech, Banking.
4 AI roles tagged guardrails.
| Company | Title | Sector | AI score | Other tags |
|---|---|---|---|---|
| Figure AI | AI Engineer, Post-Training - Helix Team | Robotics | 9 | Fine-tuning · Evals · RL post-training · Embodied AI |
| Apptronik | Senior Autonomy Software Engineer | Robotics | 8 | Embodied AI · Agent orchestration · Multimodal · Model serving · Inference infra |
| Joby Aviation | Senior AI Engineer | Robotics | 8 | Agent orchestration · Model serving · Inference infra · RAG · Vector DB · LLM observability · Evals |
| Applied Intuition | Embedded AI Engineer – Android Automotive (On-Device Intelligence) | Robotics | 7 | Model serving · Inference infra · Quantization · Fine-tuning · Tool use · Multimodal |