New AI postings mentioning Direct Preference Optimization (DPO) per week — 53 total over 11 weeks.
73 active AI roles across 29 companies mention Direct Preference Optimization (DPO). Category: ML Techniques.
73 AI roles requiring this skill.
| Company | Title | Sector | AI score | Stage |
|---|---|---|---|---|
| Anthropic | Research Engineer, Pretraining | AI Frontier | 10 | L1 |
| Writer | Staff AI research scientist | AI Frontier | 9 | L2 |
| Amazon | Director, Applied Science, Alexa for Shopping (Rufus) | Big Tech | 9 | L4 |
| Adobe | Staff Agentic ML Engineer - Photoshop | Enterprise | 9 | L4 |
| Amazon | Senior Applied Scientist, AGI Customization | Big Tech | 9 | L2 |
| Amazon | Applied Scientist, Conversational Assistant Modeling and Learning | Big Tech | 9 | L4 |
| CrowdStrike | Data Scientist, Agentic Systems (Remote) | Enterprise | 9 | L4 |
| Apple | Machine Learning Engineer | Big Tech | 9 | L5 |
| Deloitte | Research Engineer — Post-Training & Small Language Models (SLMs), Healthcare AI | Consulting | 9 | L2 |
| Amazon |
| Big Tech |
| 9 |
| L4 |
| Amazon | Applied Scientist, Selling Partner Support Engagement | Big Tech | 9 | L4 |
| Adobe | Applied Scientist 5.5 | Enterprise | 9 | L2 |
| Adobe | Applied Scientist 5 | Enterprise | 9 | L2 |
| ByteDance | Tech Lead / Principal Engineer, Creator Agent Algorithm Infrastructure | Big Tech | 9 | L4 |
| Autodesk | Principal AI Research Scientist Post-Training Alignment | Enterprise | 9 | L2 |
| Autodesk | Principal AI Research Scientist Post-Training · Alignment · Reinforcement Learning Autodesk AI Lab: London · San Francisco · Toronto · Remote (US/CA/EU | Enterprise | 9 | L2 |
| Synthesia | Principal Research Engineer | Multimodal | 9 | L1 |
| Scale AI | Research Scientist, Safety Post Training | Data AI | 9 | L2 |
| Workday | Principal AI Researcher | Enterprise | 9 | L1 |
| Autodesk | Research Lead / Principal Scientist & Manager Post-Training · Alignment · Reinforcement Learning Autodesk AI Lab: Toronto · Remote (CA) | Enterprise | 9 | L2 |
| Amazon | Applied Scientist - Agentic AI, Amazon Fulfillment Technology | Big Tech | 9 | L4 |
| Snorkel AI | AI Advocate, Open-Source & Research | Data AI | 9 | L2 |
| Cognition | Research, Post-Training | Coding AI | 9 | L2 |
| xAI | Member of Technical Staff - Post-Training and RL | AI Frontier | 9 | L2 |
| OpenAI | Researcher, Alignment Science | AI Frontier | 9 | L2 |
| Lovable | Researcher, Post Training | Coding AI | 9 | L2 |
| Autodesk | AI Research Manager/Scientist, Reinforcement Learning | Enterprise | 9 | L2 |
| Synthesia | Staff Research Engineer - Video Post Training | Multimodal | 9 | L2 |
| Amazon | Applied Scientist II, Alexa International | Big Tech | 9 | L2 |
| Amazon | Applied Scientist, Alexa Connections | Big Tech | 9 | L2 |
Direct Preference Optimization (DPO) is a skill in the "ML Techniques" category. It currently appears in 73 active AI roles across 29 companies in our index.
The top employers with active AI roles mentioning Direct Preference Optimization (DPO) are: Amazon (25), Adobe (5), ByteDance (4), Autodesk (4), Unity (3).
Over the last 11 weeks, 53 new AI postings mentioned Direct Preference Optimization (DPO). Demand is rising — up 69% in the last four weeks compared to the earliest four weeks in the window.
Roles requiring Direct Preference Optimization (DPO) are concentrated in: post-training (53%), agents (25%), pre-training (7%). These stages follow a seven-stage AI lifecycle from data preparation through to shipped product.
Job postings that mention Direct Preference Optimization (DPO) most often also require: RLHF & Alignment Techniques, Large Language Models (LLMs), Machine Learning, Reinforcement Learning (RL), Generative AI.