Software Engineer - Applied AI
$174k - $212k • San Francisco, CA • Full-time
Posted 16d ago
Job Location
San Francisco, CA
Tech Stack
Remote Work Policy
On-site
Employment Type
Full-time
Categories
AI Engineer
About the job
Deepgram is seeking an Applied AI Software Engineer to join their team in San Francisco. This role focuses on solving complex voice AI problems for enterprise customers and technology partners, often involving zero-to-one development. The team works on areas like robust ASR in noisy environments, self-service model fine-tuning and adaptation, domain-aware voice agents, agent scoring and evaluation, and on-site hardware integration. The goal is to ship production features quickly, own customer problems end-to-end, and build reusable agent modules to accelerate customer launches.
Responsibilities
- Ship production features in live voice agent products within the first quarter.
- Own customer problems end-to-end, from issue identification to production fix.
- Build reusable agent modules for ordering and reservations to speed up new customer and industry launches.
- Improve agent accuracy on real-world menus and catalogs, ensuring structured data consistency across locations.
- Contribute to fine-tuning and model adaptation for accurate customer-specific vocabulary and domains.
- Proactively identify and fix product problems or propose solutions.
- Run experiments to evaluate the impact of new functionality on customer outcomes.
Requirements
- Experience building and shipping backend services in production in the cloud.
- Proficiency in using modern AI coding tools as a primary method of work.
- Ability to learn quickly and adapt to unfamiliar parts of the technology stack.
- Proactive in identifying problems, proposing solutions, and seeing them through to completion.
- Proven experience taking a project from idea to production.
- Clear communication skills and comfort working directly with customers.
- Focus on customer outcomes with a high bar for quality and avoidance of over-engineering.
- Comfort working across multiple programming languages, primarily Kotlin and Python, with TypeScript on the front end.
