AI Research - Scientist/ Engineer

About Sentient

Sentient is building the world’s first open and community-built AGI platform — an open-source intelligence network designed to rival and complement closed systems such as OpenAI, Anthropic, and Google. Our mission is simple yet bold: to make open-source AI win.

Sentient is already a research powerhouse, delivering breakthroughs that outperform leading closed-source labs across benchmarks and applications:

• GRID – the backbone network of agents, models, and data driving cutting-edge research and powering applications.

• Sentient Chat – consumer gateway to the GRID, with 2M+ users on the waitlist.

• Dobby LLMs – open-source model family embedding human-like values at scale; ranked #1 on the UGI benchmark with the largest personality layer in AI.

• Open Deep Search (ODS) – open-source AI search framework with a 75.3% FRAMES score, surpassing GPT-4o (~65%) and Perplexity (~45%).

• ROMA – multi-agent routing framework; #1 on GitHub, outperforming Gemini 2.5 Pro and Kimi.

• Model Fingerprinting – first cryptographic framework for verifiable model ownership.

Backed by $85M in seed funding co-led by Founders Fund, Pantera Capital, and Framework Ventures, with support from Franklin Templeton, Naval Ravikant, Balaji Srinivasan, Symbolic Capital, IDG, and others, Sentient is uniquely positioned to lead the future of open-source AGI.

About the role:

As a Research Scientist/Engineer on the AI team at Sentient, you'll lead the development and implementation of techniques aimed at training language models and agents that have advanced reasoning capabilities: In particular, such AI will be able to make strategic decisions in multi-agent environments with high stakes (e.g., involving financial transactions), You'll work to develop novel post training and agentic techniques and to use these to demonstrably improve AI behavior.

Take a look at the following to get a feel for the kind of work we do at Sentient.

• Agentic frameworks: https://github.com/sentient-agi/ROMA and https://github.com/sentient-agi/OpenDeepSearch

• Post training models: https://www.alphaxiv.org/abs/2025.01v1 and https://arxiv.org/pdf/2503.16248

• Benchmarks: https://spinbench.github.io/index.html and https://arxiv.org/abs/2506.11928

Working Environment:

We are looking for multiple full time and intern positions. This is a work from home position with periodic in-person meetings in the US east coast and west coasts. You will work with other researchers at Sentient including Sewoong Oh (Professor, UW-Seattle) and Pramod Viswanath (Professor, Princeton).

Responsibilities:

• Develop and implement novel fine-tuning and reinforcement learning techniques using synthetic data generated from multi-turn interactions.

• Use these to design agentic systems to improve reasoning skills to be evaluated on long-horizon strategic decision making benchmarks.

• Create and maintain evaluation frameworks to measure reasoning skills and design new benchmarks.

You may be a good fit if you:

• Have an MS/PhD in Computer Science, ML, or related field, or equivalent experience

• Possess strong programming skills, especially in Python

• Have experience with ML model training and experimentation

• Have a track record of implementing ML research

• Demonstrate strong analytical skills for interpreting experimental results

• Have experience with ML metrics and evaluation frameworks

• Excel at turning research ideas into working code

• Can identify and resolve practical implementation challenges

Strong candidates may also have:

• Experience with language model fine-tuning and post-training

• Background in AI agents and/or reasoning research

• Published work in AI

• Experience with synthetic data generation

• Familiarity with techniques like RLHF and reward modeling

• Track record of designing and implementing novel training approaches

• Experience with model behavior evaluation and improvement

Education requirements: We require at least a masters degree in a related field or equivalent experience.