Skip to main content
Loading
loadingbar
Loading, Please wait..!!

Foundational RL Research Scientist

  • Foundational RL Research Scientist job type Posted on: Nov 10, 2026
  • Foundational RL Research Scientist experience level Brahma Consulting Group
  • Foundational RL Research Scientist employment type San Francisco, Texas
  • Foundational RL Research Scientist salary Salary: $300,000 per year
  • Foundational RL Research Scientist employment type Onsite
  • Foundational RL Research Scientist salary Full-time

Point Apply Here APPLY LATER

Curious about compensation?

Explore the historical salary trends, average pay, and estimated compensation for Foundational RL Research Scientist roles in Texas.

View Salary Guide →

Job Title :

Foundational RL Research Scientist

Job Type :

Full-Time

Job Location :

San Francisco Texas United States

Remote :

No

Jobcon Logo Job Description :

Job Description

Job Description

Brahma Consulting Group is recruiting for a venturebacked robotics company in San Francisco building learned manipulation policies for real-world deployment. They’re hiring a research scientist to own their offline reinforcement learning research.


You’ll join a small research team and lead the offline RL direction end to end. Some problems are already defined, and there’s real room to pursue your own ideas within offline RL. Success is measured by improving policy performance on internal benchmarks and extending the existing policy research. This is a research-first seat: depth in RL methods matters far more than engineering polish.


What you’ll work on

  • Developing offline RL methods that learn strong policies from large, mixed-quality robot datasets
  • Value learning, including TD-based, contrastive, and goal-conditioned approaches
  • Policy representation and how it scales with data
  • Designing rigorous experiments and benchmarks to show what actually moves policy performance
  • Turning research results into policies that run on real hardware, working with the robotics team


What they’re looking for

  • One or more first-author papers at top ML venues such as ICML, NeurIPS, ICLR, or CVPR
  • Research focused on RL methods themselves, rather than RL applied to a specific problem
  • Strong grounding in offline RL, value learning, or policy learning
  • Comfort setting research direction with a high degree of ownership


Nice to have

  • Goal-conditioned or contrastive RL research
  • Sim-to-real experience
  • Background in video or image generative models
  • Research internships at strong industry labs


Robotics experience is not required. New PhD and master’s graduates with strong publications are encouraged to apply.


Probably not the right fit if

  • Your RL experience is mainly LLM post-training (RLHF, RLVR, reward modeling)
  • Your work is primarily applied RL without methods contributions
  • You’re looking for a remote or hybrid role


Details

  • On-site in San Francisco
  • Salary range: $200,000 to $300,000+, plus equity
  • Visa sponsorship available, including new H-1B
  • Relocation support available
  • Process: short intro screen, technical conversation, technical deep dive, and on-site interview


Jobcon Logo Position Details

Posted:

Nov 10, 2026

Reference Number:

ed433a62

Employment:

Full-time

Salary:

$300,000 per year

City:

San Francisco

Job Origin:

ziprecruiter

Share this job:

  • linkedin

Jobcon Logo
A job sourcing event
In Dallas Fort Worth
Aug 19, 2017 9am-6pm
All job seekers welcome!

Foundational RL Research Scientist    Apply

Click on the below icons to share this job to Linkedin, Twitter!

Job Description

Job Description

Brahma Consulting Group is recruiting for a venturebacked robotics company in San Francisco building learned manipulation policies for real-world deployment. They’re hiring a research scientist to own their offline reinforcement learning research.


You’ll join a small research team and lead the offline RL direction end to end. Some problems are already defined, and there’s real room to pursue your own ideas within offline RL. Success is measured by improving policy performance on internal benchmarks and extending the existing policy research. This is a research-first seat: depth in RL methods matters far more than engineering polish.


What you’ll work on

  • Developing offline RL methods that learn strong policies from large, mixed-quality robot datasets
  • Value learning, including TD-based, contrastive, and goal-conditioned approaches
  • Policy representation and how it scales with data
  • Designing rigorous experiments and benchmarks to show what actually moves policy performance
  • Turning research results into policies that run on real hardware, working with the robotics team


What they’re looking for

  • One or more first-author papers at top ML venues such as ICML, NeurIPS, ICLR, or CVPR
  • Research focused on RL methods themselves, rather than RL applied to a specific problem
  • Strong grounding in offline RL, value learning, or policy learning
  • Comfort setting research direction with a high degree of ownership


Nice to have

  • Goal-conditioned or contrastive RL research
  • Sim-to-real experience
  • Background in video or image generative models
  • Research internships at strong industry labs


Robotics experience is not required. New PhD and master’s graduates with strong publications are encouraged to apply.


Probably not the right fit if

  • Your RL experience is mainly LLM post-training (RLHF, RLVR, reward modeling)
  • Your work is primarily applied RL without methods contributions
  • You’re looking for a remote or hybrid role


Details

  • On-site in San Francisco
  • Salary range: $200,000 to $300,000+, plus equity
  • Visa sponsorship available, including new H-1B
  • Relocation support available
  • Process: short intro screen, technical conversation, technical deep dive, and on-site interview


Loading
Please wait..!!