Straive is Hiring Data Scientist - Real World Data (Healthcare)

Straive; United States; Hybrid; USD 120.000 - 160.000; Vollzeit

Zusammenfassung

Straive is seeking a highly specialized Data Scientist to bridge clinical medicine and scalable data engineering in a hybrid role. You will translate complex clinical contexts into robust data pipelines, handling EHR/EMR/claims data, lab and radiology reports, and unstructured notes with NLP and LLM techniques.

You will develop prompts and pipelines to extract high-value clinical signals such as tumor-related metrics and disease progression from real-world data, working across claims, lab

Qualifikationen

• Bachelor's or Master's degree in Data Science, Computer Science, Engineering, Public Health, Biostatistics, Epidemiology, or related field.

Aufgaben

• Develop technical business logic for clinical intent on real-world data.
• Implement scalable data pipelines processing structured and unstructured data.
• Design guardrails for disease states and treatment regimens.
• Build client-specific data extracts and standardize data for analysis.
• Integrate NLP signals with structured data to enrich datasets.
• Evaluate extraction models focusing on precision and recall.
• Produce data quality reports and audits to validate logic.

Kenntnisse

SQL

Python

NLP

Data modeling

Git/version control

Clinical data knowledge

Ausbildung

Bachelor's or Master's in Data Science/CS/Engineering/Biostatistics/Epidemiology

Tools

Git

UNSTRUCTURED data handling

Jobbeschreibung

Straive is seeking a highly specialized Data Scientist to bridge the gap between complex clinical medicine and scalable data engineering. In this hybrid rolecombining the expertise of a Clinical Subject Matter Expert (SME) and a RWD Data Analyst-you will be responsible for translating intricate clinical contexts (such as rare disease diagnostic mapping, complex oncology imaging, and nuanced surgical outcomes) into robust, specification-driven data pipelines.

You will work hands-on with large healthcare datasets, translating clinical and technical requirements into reliable database queries and extract logic. By combining a deep understanding of clinical documentation (e.g., EHR, EMR, Claims data, lab and radiology reports, operative notes, unstructured EHR data) with advanced data science techniques, you will engineer prompts and pipelines that accurately extract high-value clinical signals- such as tumor volume, gradient echo sequences, and temporal disease progression from messy, real-world data.

Key Responsibilities

Clinical Rule Implementation & Technical Logic

• Develop the technical business logic that operationalizes clinical intent against messy real-world data.
• Implement clinical rules designed into scalable, reproducible data pipelines operating against the central data lake.
• Design custom data extraction guardrails for highly specific disease states, differentiating between true diagnostic characterizations and incidental findings (e.g., distinguishing post-surgical scarring from tumor recurrence).
• Develop and apply line of therapy algorithms, handling real-world complexity including combination regimens, gaps, dose modifications, and off-label use.

Pipeline Development

• Build client-specific data extracts based on well-defined technical and clinical documentation.
• Work across structured data (medical and pharmacy claims, lab results) and unstructured NLP outputs to engineer patient-level features.
• Ingest, transform, and standardize claims, EHR, lab, and registry data into analysis-ready form at scale.
• Utilize Python for handling large dataframes, finding edge cases, and processing/cleaning data using pandas.
• Follow established version-controlled workflows to support reproducibility and traceability.

NLP, LLM Integration & Prompt Engineering

• Integrate NLP-derived signals- such as diagnosis mentions, staging, biomarker results, and progression language- alongside structured claims to enrich dataset completeness.
• Analyze data and improve performance using advanced prompt engineering techniques.
• Evaluate extraction models using strict performance metrics, focusing on Precision and Recall.
• Structure and expose content so it is retrievable, semantically consistent, and reliably consumable by AI solutions and agentic workflows.

Quality Assurance & Clinical Validation

• Produce data quality reports and sample-level audits to validate clinical logic against source data.
• Perform routine review of extract outputs to ensure consistency with specifications.
• Collaborate iteratively to surface data anomalies that may indicate rule gaps, and refine logic together.

Required Qualifications

• Bachelor's or Master's degree in Data Science, Computer Science, Engineering, Public Health, Biostatistics, Epidemiology, or a related field.
• 4+ years of experience in data science, biostatistics, or health data engineering in a life sciences context.
• Strong proficiency in SQL, including complex multi-table queries, and Python for large-scale healthcare data manipulation.
• Direct experience working with claims data, EMR data, or a combination in an RWD or HEOR setting.
• Experience translating clinical or scientific documentation into database logic.
• Comfort working in governed, production-oriented data environments and collaborating with clients to understand actionable data outputs.
• Prior experience using Git or other version control systems in a collaborative environment.

Desirable Skills & Experience

• Clinical data domain expertise, familiarity with common data models such as OMOP, i2B2, PCORnet, etc..
• Exposure to Natural Language Processing outputs derived from unstructured clinical notes, pathology reports, discharge summaries, etc..
• Experience in prompt engineering involving Large Language Models (LLM) with specialized pipelines and logical engineering (e.g., handling data attrition and missingness, temporal-event logic, complex phenotyping).
• Experience working with vocabulary mapping tools and crosswalks, especially with ICD-9 or ICD-10.
• Knowledge of OpenSearch queries (good to have).
• Background in oncology or specialty disease datasets.