AI Software Validation Consultant

We are seeking an experienced AI Software Validation Engineer to support the qualification and validation of AI/ML-enabled laboratory software within a regulated pharmaceutical/GxP environment. The role will focus on the validation of AI-based automated microbiological plate counters, including IQ/OQ/PQ, AI model training and qualification, comparative performance testing, edge-case testing, data integrity assessment, and ongoing model retraining/requalification.

This is an opportunity to establish and implement an AI software validation framework from first principles, leveraging existing computer system validation (CSV) methodologies while addressing the unique risks associated with AI/ML systems. The successful candidate will work closely with Engineering, Quality, Microbiology, CSV, and CQV teams and will contribute to a broader FDA Warning Letter remediation program.

Lead and/or support the complete IQ/OQ/PQ lifecycle for AI-enabled automated plate counting systems.
Develop and execute validation strategies, protocols, test scripts, and reports for AI/ML-enabled software.
Develop AI training and qualification protocols covering both vendor-provided model training and site-specific training using client-specific plate types and labeling conventions.
Design and document a model retraining and requalification framework, including triggers, frequency, acceptance criteria, and verification requirements.
Design and execute comparative testing between AI-generated colony counts and independent human readings.
Support the implementation of the client's 200% environmental monitoring (EM) plate inspection model, involving independent AI and human reads.
Design and execute edge-case, challenge, and robustness testing for ambiguous, overlapping, atypical, or difficult-to-interpret colony presentations.
Review vendor qualification and validation documentation and identify gaps requiring additional technical information regarding the underlying AI/ML model.
Assess vendor development lifecycle controls and support or recommend supplier audits covering AI/model development, testing, governance, and change management.
Evaluate data criticality and data integrity risks for locally installed/on-premises GxP software systems.
Ensure AI-generated results that support batch release and investigations are appropriately validated, controlled, and documented.
Adapt existing organizational AI software validation and CSV lifecycle methodologies to the technical and regulatory requirements of the AI-enabled laboratory system.
Incorporate applicable regulatory and industry expectations, including relevant microbiological method validation guidance such as USP <1223>.
Monitor relevant regulatory trends and inspection findings associated with AI/ML-enabled laboratory systems.
Prepare validation documentation, risk assessments, protocols, reports, and supporting evidence suitable for FDA and regulatory inspection.
Collaborate with Engineering, Quality, Microbiology, CSV, CQV, and other cross-functional teams.
Manage validation activities effectively within an accelerated project timeline.
Provide technical guidance on AI/ML validation risks, model performance, retraining, model drift, and ongoing system governance.