SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Heidi Health is building AI-powered clinical tools that have already processed 73 million patient visits and support 2.5 million patient sessions weekly across 190+ countries. The company's mission is to double global healthcare capacity by automating administrative burden and freeing clinicians to focus on patient care.
You'll join the model team as a Senior LLMOps Engineer, responsible for building the operational infrastructure around production AI models. This is a hands-on senior role where you'll design and own systems that give the team complete visibility into model deployments, performance, and impact.
Key responsibilities include:
- Build deployment health dashboards providing live visibility into all production models, including health metrics, monitoring, and proactive incident alerting
- Create complete session-to-model lineage tracing so degradations can be diagnosed from affected sessions to root cause in minutes
- Establish a feedback flywheel that converts Intercom tickets and CSAT feedback into training data via AI agents
- Retrieve and summarize complete execution traces for flagged sessions for model team review
- Own per-model P&L analysis, measuring revenue against inference costs to inform deployment strategy
- Implement LLMOps best practices (tracing, evaluation, model incident response) across the organization
- Partner with researchers and engineers across ASR, note generation, Evidence, and Dictate models
You'll need 2–3 years of hands-on LLMOps experience at a mature AI company (preferably US or China-based), with proven ability to ship monitoring systems (Datadog or similar), distributed tracing across multi-step LLM pipelines, and data models that scale. You should have experience building with LLMs (not just operating them), comfort with unit economics analysis, and senior-level ownership mentality. A broader engineering foundation (backend, data platform, ML infrastructure) is a plus, as is experience with product feedback tools like Intercom or background in regulated/safety-critical domains.
