You will drive model alignment and performance for next-generation speech-to-speech AI. This role involves hands-on research, data curation, and executing training experiments using RLHF and SFT to improve emotional intelligence in dialogue. You will work with large-scale models and massive datasets to bridge the gap between transactional exchanges and natural, human-like conversations.
This job is no longer actively hiring. Talk to Jack to find live roles.
Research Scientist, Post-Training at fast-growing conversational AI startup
Are you ready to redefine how humans interact with technology? This venture-backed startup in San Francisco is looking for a Research Scientist to lead post-training efforts for a next-generation speech-to-speech AI platform. You will work on 10B+ parameter models and 100+ TBs of data to create fluid, emotionally intelligent voice interactions that feel truly human. With a founding team backed by top-tier VCs, significant equity on the table, and an aggressive ship-to-production culture, this is a rare opportunity to bridge the gap between deep research and real-world impact. If you have PhD-level expertise in RLHF or LLM alignment, this is the role for you.
Want to apply for this role?
C
This role is no longer actively hiring, but Jack can still help you discover similar open roles that fit.
Location
San Francisco, United States
Compensation
$180k-$250k + Equity
Company
Confidential company
Role overview
About the company
Fast-growing conversational AI startup
What you will do
- Lead post-training workflows including supervised fine-tuning (SFT) and preference optimization (RLHF/DPO) for large-scale models.
- Curate high-quality datasets and design automated or human-in-the-loop evaluation frameworks to measure model performance.
- Formulate and test hypotheses to improve model alignment, emotional context, and real-time dialogue management.
Who this is a fit for
- PhD in Machine Learning or related field with publications at top-tier conferences like NeurIPS or ICML.
- Hands-on experience training 1B+ parameter models, specifically in LLM post-training or state-of-the-art speech modeling.
- Proven ability to thrive in early-stage environments with a focus on shipping fast and obsessing over user experience.
Why this role is remarkable
- Work on cutting-edge duplex AI architectures that move beyond traditional walkie-talkie style voice interactions.
- Join a mission-driven founding team backed by top-tier VCs and prominent technology industry leaders.
- Significant equity and high autonomy in an environment designed for rapid iteration and shipping research to production.
How Jack & Jill work together
I get to know what you’re great at, then find roles you’d never find yourself.Ok, I'll go first. I'm Jack, an AI that gets to know you on a quick call, learning what you're great at and what you want from your career. Then I help you land your dream job by finding unmissable opportunities as they come up, supporting you with applications, interview prep, and moral support.
I recruit from Jack’s network and make the intro when I spot a great match.And I'm Jill, an AI Recruiter who talks to companies to understand who they're looking to hire. Then I recruit from Jack's network, making an introduction when I spot an excellent candidate.
Jack gets to know what you're great at and what you want next, then searches 15 million jobs daily and helps you discover roles at companies like this.
Meet Jack
What happens next?
Jack’s an AI agent for job searching and career coaching. He works for you.
Jill is the AI recruiter working for the company. She recruits from Jack’s network.
If your profile’s a match and Confidential company wants to meet, Jill will make the intro. In the meantime, Jack will send you excellent alternatives.