Clinical Data SME (Biometrics & Data Engineering)

Role Overview

This role is the connective tissue between Biometrics and Data Engineering. It exists because clinical data has meaning that structure alone doesn’t carry: someone has to understand what the data represents clinically, what the standards require of it, and how it must therefore be modeled — and then make sure the engineering team builds systems that produce it correctly.

The SME owns that translation. They bring clinical data domain expertise and a solid working command of CDISC standards, and they use it to guide the design of data products that are clinically correct, standards-conformant, traceable, and ready to support analysis and an AI-ready data strategy.

Working knowledge of CDISC, SDTM, and ADaM is mandatory — specifically, at the level required to design systems and pipelines that enforce the standards, map source data into them, and extend them where the data demands it.

The CDISC Expertise This Role Requires

This is worth being precise about, because it defines who fits.

What we need: someone who can look at source data and design the system that gets it correctly into SDTM — mapping source to target domains, enforcing controlled terminology through validation rather than manual review, and extending or customizing domains where the data genuinely requires it. This is applied, systems-oriented CDISC expertise: knowing the standard well enough to build against it and to tell an engineering team what “conformant” actually means in code.

What we don’t need: someone to author the standards themselves. This role doesn’t define new controlled terminology, sit on standards bodies, or own submission deliverables. It consumes the published standards and makes sure our systems honor them.

What This Person Actually Does

Owns the clinical meaning of the data. Serves as the authority on what the data represents — how it’s collected, what the clinical workflows behind it imply, and what constraints that places on how it can be structured. This is the judgment that engineering teams cannot supply for themselves.

Designs how the standards get enforced. Specifies how source and EDC data maps into SDTM target domains, and how controlled terminology is validated in the pipeline rather than checked after the fact. Determines when data fits an existing domain, when supplemental qualifiers suffice, and when a domain needs to be extended or customized to represent the data honestly — then makes sure that logic is built consistently rather than reinvented per study.

Translates in both directions. Turns clinical and business needs into requirements, mapping specifications, and user stories engineering can act on — and explains technical constraints and trade-offs back to Biometrics stakeholders in terms they can decide on.

Protects traceability. Ensures an unbroken path from source data through SDTM to analysis, so any value can be explained and defended. In a regulated environment this isn’t documentation hygiene; it’s what makes the data defensible.

Integrates the data that doesn’t arrive neatly. Brings external sources — labs, PK, biomarkers, imaging, eCOA — into standardized structures, resolving the mismatches in keys, terminology, and timing these sources invariably introduce.

Guides delivery without owning the build. Works alongside data engineers and architects, supplying clinical and standards input, reviewing designs, and validating that what’s delivered means what it’s supposed to mean. Engineering owns the how; this role owns the what and the why.

Closes the loop on adoption. Supports validation and testing against real clinical scenarios rather than functional checks alone, and stays engaged until the people who asked for the data product are actually using it.

Sets the direction. Drives clinical data strategy by embedding CDISC standards, data domains, and governance into how data is built — the discipline that makes datasets trustworthy enough to be analytics- and AI-ready.

What We’re Looking For

Mandatory — CDISC / SDTM / ADaM (applied)

• Solid working knowledge of CDISC, sufficient to design systems and pipelines that produce conformant data.
• Demonstrated source-to-SDTM mapping: the ability to take raw, EDC, or legacy clinical data and map it correctly into target SDTM domains.
• Practical command of controlled terminology — enough to design validation that enforces it systematically, not to author it.
• Ability to extend or customize SDTM domains where the data requires it, with the judgment to know when that’s warranted versus when existing domains or supplemental qualifiers are the right answer.
• Working understanding of ADaM and how analysis-ready data derives from SDTM, sufficient to guide design and preserve traceability.
• Familiarity with what conformance and validation expectations mean in practice, and with the SDTM Implementation Guide as a working reference.

Also Required

• Clinical data domain expertise — a real understanding of how clinical data is collected and why that shapes how it must be modeled.
• Hands-on familiarity with EDC systems and the realities of clinical data collection.
• A demonstrated ability to gather clinical and business requirements and convert them into technical specifications and user stories engineering teams can build from.
• Experience partnering with data engineering and architecture teams in a supporting, advisory capacity — comfortable guiding design without taking over implementation.
• Working knowledge of clinical data workflows, external data integration, data standards, and data governance in a regulated (GxP / 21 CFR Part 11) environment.
• Strong communication and stakeholder management skills, and fluency in both directions — able to hold a conversation with Biostatistics, Statistical Programming, Clinical Operations, and Regulatory on one side, and data engineers and architects on the other.

Education & Qualifications

Required

• Bachelor’s degree in Computer Science, Information Systems, Data Science, Software Engineering, Statistics, Applied Mathematics, Bioinformatics/Health Informatics, or a related technical or quantitative discipline.
• Equivalent demonstrated experience in data architecture, data modeling, or data engineering will be considered in place of a formal degree. What matters is the ability to reason rigorously about data structure, relationships, and integrity — however that capability was acquired.

Preferred

• Master’s degree in a data-oriented field — Data Science, Analytics, Computer Science, Information Systems, Biomedical/Health Informatics, or Biostatistics.
• Formal coursework or credentials in data modeling, data architecture, or database design.
• Certification in a relevant data discipline:
• Data management or governance (e.g., CDMP / DAMA-certified data management professional)
• Data architecture or solution architecture
• Cloud data platforms (e.g., AWS, Azure, or equivalent data-focused certification)
• Business analysis (e.g., CBAP/CCBA) or an equivalent requirements-engineering credential
• CDISC training or certification.

A Note on Background

We are deliberately open on where the domain knowledge came from. Candidates from a technology, data, or informatics background who have since built genuine working depth in clinical data and CDISC are strongly encouraged to apply — this role is designed for exactly that intersection.

What is not negotiable is that the CDISC knowledge be real and applied: someone who can map source data to SDTM and design systems that enforce the standard, not someone who has only read about it.

What Success Looks Like

Clinical data that is consistently mapped and conformant to CDISC across studies, traceable end to end from source through SDTM to analysis, and trusted by the people who use it — with engineering teams building the right thing the first time because the requirements and mapping logic were clear, and with Biometrics stakeholders confident that the data products they receive reflect what they actually meant.