> Markdown version of [/jobs/ext/3756679-lead-data-scientist-data-analyst](https://www.wearedevelopers.com/jobs/ext/3756679-lead-data-scientist-data-analyst). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Lead Data Scientist / Data Analyst

- **Company:** EPAM Systems, Inc.
- **Location:** Newtown, PA, United States (Remote available)
- **Experience:** Expert
- **Contract:** Permanent contract
- **Skills:** Computer Vision, Encodings, Machine Learning, Natural Language Processing, Operational Databases, Feature Engineering, Evaluation Pipelines, Machine Learning Operations, Software Version Control, Databricks
- **Published:** October 15, 2026
- **Apply:** https://arc.dev/remote-jobs/j/redirect/prgocnqu8d

## About the Role

* 5+ years of experience in data science and machine learning for classification problems
 * Hands-on experience building hybrid deterministic + ML solutions for ambiguous cases
 * Lead-level ownership of end-to-end model delivery from prototyping through stabilization and knowledge transfer
 * Strong project execution skills in fixed-scope, milestone-driven delivery with multiple parallel model tracks
 * Hands-on Python.ML skills for feature engineering, training, and evaluation workflows
 * Databricks and MLflow experience for reproducible training, evaluation, and model versioning
 * Computer vision and natural language processing experience to support fuzzy matching and embedding-based approaches
 * OMOP Common Data Model knowledge and SNOMED CT Veterinary Extension concept-mapping experience
 * Strong analytical skills to define class-frequency tiers, calibrated risk bands, and performance thresholds
 * Strong communication skills to interpret annotation and validation plans with domain SMEs and translate them into model updates
 * Advanced English proficiency (C1+ level)

Nice to have

 * Expertise with classification metrics for imbalanced data such as PR AUC, F1 by class, and balanced accuracy
 * Experience using calibrated probability and reliability diagrams to assess confidence scores
 * Ability to set per-tier metric targets and acceptance criteria for rare and clinically significant classes

## Description

* Lead design and delivery of optimized models for species, gender, breed, weight classification, and age classification
 * Build health-predisposition insights models using standardized signalment attributes and governed ontologies
 * Apply a deterministic-first approach and use machine learning primarily for high-ambiguity scenarios
 * Define class-frequency tiers, calibrated thresholds, and evaluation plans that go beyond overall accuracy
 * Produce confidence scores, method traces, exception flags, and model versioning artifacts for each model
 * Create model cards and documentation that support governance, reproducibility, and transition to operations
 * Validate performance across common, rare, ambiguous, and clinically significant cases using appropriate classification metrics
 * Incorporate animal-health domain knowledge into life-stage thresholds, breed-dependent reference ranges, and risk mappings
 * Implement training and evaluation pipelines in Databricks with MLflow tracking and reproducible builds
 * Collaborate with SMEs to refine labels, interpret validation feedback, and adjust modeling strategies
 * Support integration into production data pipelines and contribute to stabilization and knowledge transfer