Machine Learning Engineer

Role: Machine Learning Engineer

Kuddo Health is hiring its first ML engineer to lead end-to-end modeling work alongside the CTO and Stanford Medicine researchers. You’ll translate clinical fidelity rubrics into computational tasks, build and evaluate labeled data pipelines, fine-tune open-weight models as data grows, and ship an on-prem research pipeline.

What you’ll do

Clinical fidelity → ML problem

• Translate clinical fidelity rubrics into a well-posed extraction-and-scoring task
• Own the architecture of the fidelity scoring engine
• Reason about model behavior (what it should learn, where it fails, and what evidence proves it works)

Data and ground truth

• Curate and preprocess expert-scored therapy session data
• Design labeling workflows with clinical reviewers
• Handle messy inputs (e.g., differences between clean Zoom recordings and noisier archival audio)
• Build reproducible data infrastructure and contribute to onboarding + ETL/ELT

Evaluation against expert raters

• Create and document test plans
• Run evaluations including: kappa, percent agreement, calibration, and failure-mode analysis
• Use disagreement cases as diagnostic signal to guide next training iterations
• Maintain robust documentation

Fine-tuning and model iteration

• Fine-tune open-weight models as more expert-scored sessions accumulate (e.g., LoRA, instruction tuning, or full fine-tuning)
• Run experiments, track results, and decide what ships with documentation

On-prem research deployment

• Ship the pipeline into an air-gapped/on-prem research cluster with no internet or public-facing networks
• Design for constraints from the start
• Partner with a fractional CTO on cybersecurity/IT security decisions
• Create test plans to identify the application’s failure modes/risks (as described in the posting)

Working with leadership

• Reports to the CEO
• Works directly with the CTO and Stanford Medicine researchers

Contract / timeline

• Contract for the first three months (paid from the research project budget)
• Full-time conversion after the company closes its seed round this summer

Requirements

• The posting does not list explicit minimum requirements; the role emphasizes end-to-end ownership of ML modeling, data pipeline building, evaluation, fine-tuning, and on-prem deployment.