Senior Machine Learning/Data Engineer
Senior ML/Data Engineer at mental-health provider PsychPlus builds, deploys and owns ML models on clinical and operational data end to end — training, production serving APIs, monitoring/retraining — plus defines core business metrics in the lakehouse. Core stack: Python, PySpark, SQL, and Azure/Microsoft Fabric (or Databricks/AWS/GCP).
- Compensation
- Not disclosed
- Location
- Sindh, Pakistan
- Work setup
- Not specified
- Seniority
- Senior
- Employment
- Not specified
- Experience
- 3+ years
- Category
- Data engineering
- Posted
Skills
Must have
- Fluent in English, written and spoken; daily work with colleagues in Europe and the United States
- 6+ years building machine learning models professionally
- 3+ years of models personally deployed and operated in production on live data
- Strong understanding of gradient boosting and tree-based methods, feature engineering on tabular/temporal data, class imbalance, probability calibration, threshold selection
- Expert-level Python for production systems (modular, tested, version-controlled code)
- Advanced PySpark
- Expert SQL including complex joins, window functions and query optimisation over large schemas
- Proven experience building and deploying backend services or APIs (REST or GraphQL) that serve predictions in production
- 3+ years on a modern cloud data platform; Microsoft Fabric or Azure strongly preferred
- Hands-on MLOps: experiment tracking, model versioning, reproducible training pipelines, production monitoring (e.g. MLflow)
- Experience with real-world messy data: reverse-engineering source systems, finding silent data-quality failures
Nice to have
- Healthcare data experience (EHR, claims, billing, clinical records) and work under HIPAA or comparable
- Dashboarding with Power BI or similar, including semantic model ownership
- Experience with survival analysis, uplift modelling, causal inference or optimisation under constraints
- Deep-learning experience welcome
Full description
The current delivery model for mental health care is broken in this country. PsychPlus set out on a mission to reimagine how mental health care is delivered to folks who need it. Our goal is to provide a digital-first, modern approach to psychiatry and therapy. We provide care both in-person at our offices across Texas and virtually so our patients can be seen at their convenience.
Through a combination of exceptional medical and psychological care and best-in-class technology, we provide an unparalleled approach that serves our patients' needs in an integrated way. Join us in our mission to ensure that every person has access to affordable and accessible mental health care.
About The Role
We are looking for a Senior Machine Learning/Data Engineer to join our Data/AI/ML team. This is a hands-on, end-to-end role; you will build machine learning models, put them into production, and be responsible for them once they are running.
The role deliberately spans three areas that are often separated. You will do applied machine learning (build and evaluate models on real clinical and operational data), production engineering (deploy, serve and monitor them reliably), and data and analytics (understand messy source data, define the metrics the business runs on, and turn analysis into decisions).
You will work directly with the AI Team and closely with our data platform, engineering, clinical and revenue-cycle teams. Because the team is small, your work has immediate visibility; models you build are discussed at executive level within weeks of going live.
Examples of the projects that you will work on include (but not limited to): appointment attendance prediction, model performance and iteration, revenue-cycle analytics, prediction of mental health states, core business metrics, etc.
Responsibilities
- Build, evaluate, and ship models. Develop ML models on clinical, and operational data, and take them all the way to production, including feature pipelines, training, evaluation, deployment and rollout.
- Build the monitoring, alerting and retraining processes that keep models healthy. Detect drift and degradation, and diagnose root causes when a model is not delivering.
- Build and maintain the backend services and APIs through which our EHR consumes model predictions. Write production-grade, tested, version-controlled code — batch and request-response.
- Understand and shape the data: Go deep into our EHR and claims systems to work out what the data actually means, where it is wrong, and how to make it usable. Document what you learn so the next model is faster to build.
- Design and implement the metrics that the AI team as well as the company relies on, with defensible definitions, and productionize them as scheduled jobs in the lakehouse.