Staff ML Engineer, Frontier AI

present of AI; San Francisco (CA); Sur place; USD 180 000 - 260 000; Plein temps

Résumé du poste

Ambience Healthcare in San Francisco is seeking a senior AI/ML engineer to shape the AI behind our clinical documentation and coding products. This hands-on role requires deep experience with production-grade large language models, robust evaluation design, and agentic systems, alongside strong Python and ML engineering skills.

You will lead cross-functional teams, translate vague quality problems into measurable tests, and contribute code while digging into failures to improve system

Qualifications

• Deep experience with large language models in production.
• Strong evaluation design and measurement.
• Experience with agentic systems and post-training methods.
• Solid Python and ML engineering fundamentals.
• Experience shipping consequential AI systems.

Responsabilités

• Set direction for AI behind clinical documentation and coding products.
• Lead across teams while writing code and debugging failures.
• Turn vague quality problems into measurable tests.
• Design and improve evaluation pipelines and feedback loops.

Connaissances

Production LLMs

Evaluation design

Agentic systems

Python ML engineering

Outils

Python

PyTorch

Description du poste

AI & Data · San Francisco · FullTime · Senior level

This is a hands on senior technical role setting direction for the AI behind clinical documentation and coding products. It asks for deep experience with large language models in production, strong evaluation design, agentic systems, and post training methods, plus solid Python and ML engineering. A strong candidate has shipped consequential AI systems, can turn vague quality problems into measurable tests, and can lead across teams while still writing code and digging into failures.

How it fits the AI era

The work is already built on AI, so the role is about making AI systems reliable rather than introducing them. Much of the judgment stays human: deciding which failures matter, designing fair evaluations, and choosing when to retrain a model versus adjust prompts, retrieval or surrounding systems. Clinical context and accountability for real patient facing outcomes remain with people.

How it builds AI experience

Someone here would gain hands on experience improving models for clinical documentation and coding, including post training, evaluation pipelines, and feedback loops built from real clinician edits. Building that background means practicing evaluation design, failure analysis, and post training on messy real world data, ideally with speech or multimodal systems. Healthcare or other regulated settings would add useful context.

Worth showing

• Production LLM systems
• Agentic system architecture
• Post training methods
• Python and PyTorch

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Ambience Healthcare

AI that listens to clinical visits and drafts the documentation, coding and clinical checks so doctors spend less time on the screen.