
Deepgram
Research Engineer, Post-Training at Deepgram
$270,000–$338,000 / year
Apply for this positionRole overview
An at-a-glance view of the facts included with this Research Engineer, Post-Training at Deepgram listing.
- Company
- Deepgram
- Location
- Remote, United States
- Work arrangement
- Remote
- Employment type
- Full-time
- Compensation
- $270,000–$338,000 / year
Applications for this listing are submitted through CareersForge. Use the form on this page, then sign in to your Candidate Portal to follow application updates and any interview invitation.
Role overview
Deepgram is hiring a full-time Research Engineer, Post-Training, for a remote role in the United States. The listed compensation is $270,000–$338,000 per year. The work focuses on finding shortcomings in voice AI models, defining measurable improvement goals, and delivering fixes to production—often within weeks.
Role at a glance: Remote U.S. full-time role improving and shipping Deepgram voice AI models, with listed pay of $270,000–$338,000 per year.
Responsibilities mentioned in the listing
- Identify model limitations and establish targets for improvement.
- Create evaluations that show whether models improve and explain why.
- Curate and generate data, then fine-tune models.
- Build working systems quickly and release them to production.
- Take customer problems from initial measurement through verified improvement in production.
- Find practical routes to solutions, including fine-tuning, new data, or external models.
- Develop capabilities such as support for additional languages and regions, clearer pronunciation of names and numbers, and speaker isolation amid noise or crosstalk.
Qualifications stated in the listing
- Deep experience training, fine-tuning, and evaluating deep learning models.
- Strong Python and PyTorch skills.
- A record of deploying machine learning systems in production.
- Ability to own problems end to end and learn new skills as needed.
- Sound judgment about which problems matter and when results are ready for customers.
- No specific degree is required; relevant evidence may come from challenging projects, shipped systems, or open-source work.
- Additional strengths may include experience with speech-to-text, text-to-speech, or voice agents; distributed training; and published research, open-source contributions, or public benchmarks.
- Ability to build and verify AI workflows, take responsibility for their outputs, adapt to changing tools, and apply personal judgment about AI use.



