16 Nov
|
HTC Global Services
|
Chennai
16 Nov
HTC Global Services
Chennai
Day to day
- Design, train and test predictive models on patient data over time.
- Compare models against baselines and benchmarks agreed with medical and client teams.
- Explain model results in plain terms for medical reviewers.
- Package tested models so engineers can plug them into the product, then monitor them.
Firm customer requirements
- MUST have personally built healthcare predictive models on longitudinal claims, Real World Evidence (RWE) or clinical data.
- MUST have robust statistics and machine-learning skills: model selection, calibration, discrimination, error analysis and uncertainty.
- MUST have time-to-event or longitudinal modeling experience, including feature engineering, class imbalance, missing data and preventing time leakage.
- MUST use explainability methods suited to medical audiences.
- MUST be able to set benchmarks, baselines,
validation datasets and release criteria with medical and Client stakeholders.
- MUST write production-grade Python with testing, documentation and reproducibility, and work with MLOps or software engineers.
- Strongly preferred: real-time or near-real-time model serving, latency limits, model monitoring and drift checks.
What they deliver
- Validated models with documented assumptions and limits.
- Benchmark comparisons, sensitivity analyses and performance reports.
- Explainability outputs ready for medical review.
- Tested model files and interfaces ready to integrate.
- A plan for monitoring and revalidating models once in use.
📌 Data Engineer (Chennai)
🏢 HTC Global Services
📍 Chennai