Machine Learning Engineer

Hyperbound•San Francisco, CA

5d•$260,000 - $300,000•Onsite

About The Position

Hyperbound is seeking a Machine Learning Engineer to build and ship the models that power their roleplay, scoring, and coaching products. This role involves owning the full lifecycle of models, from training and fine-tuning to deployment and maintenance in production. The engineer will work with open-source models, deploy them on-device to meet customer latency and privacy requirements, and build robust evaluation frameworks and benchmarks to ensure model performance. This is a challenging but exciting role where the engineer will have significant input into product development and work closely with founders and the engineering team.

Requirements

Own models end to end, from training through production.
Spend most of the time fine-tuning and running open-source models in production.
Push models on-device.
Build evaluation frameworks that determine if a change actually worked.
Work hard and have high standards.
Be a builder first.
Not be precious about code and fix things that are broken.
Be comfortable with the tradeoffs around quantization and distillation.

Nice To Haves

Experience with sales technology.
Experience with agentic operating systems.
Experience with revenue activation platforms.

Responsibilities

Build and ship models for Hyperbound's roleplay, scoring, and coaching products.
Own the full lifecycle of models, including training, fine-tuning, deployment, and ongoing maintenance in production.
Fine-tune and deploy open-source models for greater control over cost, latency, and functionality.
Deploy models on-device to meet customer latency or privacy requirements, involving quantization and distillation.
Build evaluation frameworks, benchmarks, and regression suites to assess model performance.
Collaborate closely with founders and the engineering team, providing input on product development.
Work in the office five days a week in San Francisco.