Machine Learning Engineer, Data Products
Description
About mpathic
mpathic builds AI that understands clinical and human conversations: analyzing therapy and clinical-trial sessions, producing expert-labeled data for AI builders, and publishing open safety benchmarks.
About the role
We’re looking for a Machine Learning Engineer to build and improve ML-powered products, with a particular focus on the data that makes them effective. You’ll work across the ML lifecycle, combining hands-on engineering with strong data skills to create reliable datasets, pipelines, and data-driven product capabilities.
This is a generalist ML engineering role with a data focus. You’ll contribute to production systems while improving how data is collected, prepared, accessed, and used for training and inference.
What you’ll do
- Build, deploy, and maintain ML models and services in collaboration with engineering and product teams.
- Build reliable ML pipelines from the ground up for conversational audio and transcripts: ingestion, validation, diarization, and overall quality.
- Create and maintain versioned datasets and features for training, evaluation, and inference.
- Provision and manage AWS resources your pipelines and tools run on, using infrastructure as code.
- Establish checks for data quality, freshness, schema changes, and training-serving consistency.
- Build workflows for labeling, enrichment, deduplication, and dataset curation with our clinical experts.
- Explore data to identify quality issues, coverage gaps, and opportunities to improve model performance.
- Own well-scoped projects and communicate data assumptions, limitations, and trade-offs clearly.
What you’ll bring
- Practical experience shipping ML systems, typically 2–6 years of relevant work or equivalent.
- Strong Python skills, with experience building and operating real-world data systems at scale, including storage, serving, dealing with contamination, noise and real-world data quality challenges.
- Proven track record of excellence building and maintaining data pipelines and reproducible, versionable workflows.
- Solid ML fundamentals and an understanding of how data quality affects model performance.
- Hands-on AWS experience (EC2, S3), infrastructure as code (Terraform or CDK) plus at least one database or warehouse.
- Strong experience processing and manipulating datasets with inconsistent schemas, and imbalanced or unrepresentative data.
- Familiarity with data modeling, storage formats, and batch or streaming processing.
- Experience standing up processes and systems from the ground up, and sound judgement with handling sensitive health data.
- Good software engineering practice: testing, version control and code review.
Helpful additional experience
- Workflow orchestration and distributed processing tools, such as Airflow, Dagster, or Spark.
- Docker and CI/CD.
- Dataset versioning, lineage, and reproducibility.
- Annotation platforms, active learning, or human-in-the-loop workflows.
- Comfortable shipping quickly and iterating on feedback.
- Building data APIs or internal tools with lightweight frameworks (FastAPI, Streamlit, etc.).
- Clear, direct communication, both written and verbal.
- Startups and cross-functional teams.