SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Sierra is a leading platform for customer-facing AI agents serving major brands including The GAP, Rocket Mortgage, SoFi, Sutter Health, and SoftBank. The company is headquartered in San Francisco with offices across North America, Europe, and Asia.
The Applied Research team advances the quality, speed, and cost efficiency of models powering Sierra's agents. This role focuses on solving applied AI and ML problems spanning post-training of open-weight models, voice AI, and long-horizon task execution.
You will tackle unsolved applied AI and ML problems, pushing boundaries in post-training, voice AI, and complex task execution. Key responsibilities include: determining when to build specialized models versus leveraging external frontier models while balancing quality, speed, and cost; translating ambiguous product and customer problems into well-framed ML problems; developing, fine-tuning, evaluating, and deploying models that improve customer-facing agents; advancing post-training work from supervised fine-tuning toward reinforcement learning and long-horizon optimization; building rigorous evaluations connecting model performance to real-world agent quality; working with large-scale proprietary datasets spanning text and audio; and collaborating across the company to move ideas from experimentation through production deployment.
You will operate with high agency, work across the stack when needed, and demonstrate strong judgment in selecting the simplest and most effective approaches. The role requires excellent judgment in framing product problems as ML tasks, practical evidence-driven thinking about when modeling is needed versus simpler solutions, and deep experience with data, experimentation, evaluation, and model iteration.
REQUIREMENTS:
- Strong experience building and deploying production machine learning and AI systems (backgrounds in applied ML including search, ranking, recommendations, advertising, language, and speech welcome)
- Excellent judgment in framing product problems as ML tasks and selecting the simplest and most effective approach
- Deep experience working with data, experimentation, evaluation, and model iteration
- Practical, evidence-driven approach to ML with strong judgment about when modeling is needed
- Ability to operate with high agency and work across the stack to deliver outcomes
- Degree in computer science, machine learning, or related field, or equivalent professional experience
- Alignment with company values: Craftsmanship, Customer Obsession, and Competitive Intensity
PREFERRED QUALIFICATIONS:
- Experience with supervised fine-tuning or reinforcement learning
- Experience in search, ranking, recommendations, or advertising systems
- Experience training and deploying language, speech, or multimodal models
- Experience designing model evaluations tied to real-world product outcomes
- Experience working with large-scale consumer datasets or high-volume ML products
- Contributions to applied ML research or benchmarks
