Runs the team day to day
In-house, owns the opening
Joined the team recently
From your own Gmail and LinkedIn.
About the role
This is a full-time senior data scientist role in Noida, India, for someone with four to eight years of production ML or data science experience in healthcare data. The role leads a small data science team while staying hands-on, shaping modeling approaches, reviewing code and experiments, and moving models into production. The main technical focus is probabilistic entity linking and deduplication across large, inconsistent healthcare datasets, using graph methods. Doceree builds AI tools for healthcare marketing that connect messages to healthcare professionals in a privacy-compliant way.
What you’ll do
- Extract entities and attributes from warehouse tables, registries and data feeds.
- Design and build probabilistic entity linking and deduplication across messy healthcare data.
- Own the linking pipeline, from pairwise and graph-based scoring through calibrated match probabilities to stable entities.
- Define link quality measures such as precision, recall, calibration and confidence thresholds, and run error analysis.
- Map entities and attributes to controlled medical vocabularies for specialties, conditions and products.
- Apply graph algorithms and graph ML, including clustering, centrality, community detection, link prediction and embeddings.
- Build and maintain the data and graph platform, including warehouse transformation, monitoring and performance tuning.
- Lead and develop a small team of data scientists and help improve how the team operates.
What they’re looking for
- Four or more years shipping production ML or data science work, including hands-on entity linking, resolution or record linkage at scale.
- Production-grade Python with tests and type hints, plus strong SQL.
- Software engineering practice with FastAPI, Docker, pytest, Git and CI/CD.
- Probabilistic methods such as match scoring, calibration, threshold selection and evaluation.
- Graph algorithms such as connected components and centrality, plus graph ML such as embeddings or GNNs.
- Warehouse-centric pipelines, ETL, scalable blocking, vector search and incremental refresh.
- Experience with Snowflake or a similar distributed SQL warehouse, PyTorch Geometric or DGL, igraph or NetworkX, Splink, and a graph store such as Neptune or Neo4j.
Nice to have
- Knowledge of medical ontologies such as MeSH, UMLS, SNOMED or RxNorm.
- Experience with prescribing or claims data, NPPES or NPI, PubMed or ClinicalTrials.gov.
- Biomedical NLP, AdTech experience, or work in regulated environments with PII constraints.
Skills
Questions about this role
Where is the job located?
Noida, India.
How much experience is needed?
Four to eight years, with at least four years of production ML or data science work including entity linking, resolution or record linkage.
Does the role involve managing people?
Yes. It leads a small data science team while also contributing technically.
What tools should I know?
Python, strong SQL, Snowflake or a similar warehouse, graph ML libraries, Splink, and a graph database such as Neo4j or Neptune.