Senior AI/ML Engineer, Production AI (Contractor)
Legend Biotech
4 months ago
Remote friendly (Somerset, NJ)
United States
IT
Senior AI/ML Engineer, Production AI (Contractor)
Role Overview
- Develop, design, deploy, monitor, and govern enterprise-ready ML and Generative AI systems that are scalable, auditable, and compliant with internal AI policies and regulatory expectations.
- Help establish MLOps and GenAI Ops foundations, including evaluation, observability, and Responsible AI controls.
Key Responsibilities
- AI/ML & GenAI Engineering
- Design, build, and deploy production-grade ML and Generative AI solutions from prototypes to hardened services.
- Implement GenAI patterns: RAG; prompt engineering and prompt versioning; embedding pipelines and vector search; secure API-based model access.
- Ensure AI systems meet enterprise standards for scalability, performance, reliability, and security.
- MLOps & GenAI Ops Frameworks
- Build/configure end-to-end MLOps and GenAI Ops frameworks: model and prompt versioning; reproducible pipelines and CI/CD; controlled deployment and rollback strategies.
- Integrate AI workflows with enterprise data platforms, orchestration tools, and cloud infrastructure.
- Model & GenAI Evaluation
- Define evaluation frameworks for ML and GenAI (accuracy/robustness/drift; LLM response quality, grounding, hallucination risk, safety checks; bias/fairness/explainability).
- Establish acceptance criteria and validation artifacts for regulated, audit-ready environments.
- Observability & Monitoring
- Implement observability to monitor ML/LLM degradation, data/embedding drift, prompt/response behavior, latency/failure modes/usage patterns.
- Enable full logging and traceability for investigations, audits, and continuous improvement.
- Responsible & Ethical AI
- Apply Responsible AI principles (human-in-the-loop controls; transparency/explainability/proper-use disclosures; privacy/access control/lineage).
- Ensure GenAI features are opt-in, governed, and aligned with AI policies and regulatory expectations.
- Collaboration & Leadership
- Partner with Data Engineering, Architecture, Security, QA, and Business teams.
- Translate business problems into well-scoped, governed AI/GenAI solutions.
- Contribute to enterprise AI standards, reference architectures, and platform roadmaps.
Requirements
- Bachelor’s or Master’s degree in Computer Science, Engineering, Data Science, or related field.
- 5+ years deploying ML systems in production.
- Strong experience with: Python; ML frameworks (PyTorch, TensorFlow, scikit-learn); LLMs/GenAI tooling; MLOps practices (pipelines/automation); cloud platforms (Azure/AWS/GCP).
- Familiarity with vector databases, embedding strategies, and RAG & graph architectures.
- Proven ability to design governed, observable, and secure AI systems.
- Experience in biotech/life sciences/healthcare or other GxP-relevant domains.
- Extensive experience with enterprise SDLC and production IT processes; full SDLC delivery of AI systems.
- Experience implementing GenAI in enterprise or regulated environments.
- Exposure to AI governance, risk assessments, or validation frameworks.
Role Overview
- Develop, design, deploy, monitor, and govern enterprise-ready ML and Generative AI systems that are scalable, auditable, and compliant with internal AI policies and regulatory expectations.
- Help establish MLOps and GenAI Ops foundations, including evaluation, observability, and Responsible AI controls.
Key Responsibilities
- AI/ML & GenAI Engineering
- Design, build, and deploy production-grade ML and Generative AI solutions from prototypes to hardened services.
- Implement GenAI patterns: RAG; prompt engineering and prompt versioning; embedding pipelines and vector search; secure API-based model access.
- Ensure AI systems meet enterprise standards for scalability, performance, reliability, and security.
- MLOps & GenAI Ops Frameworks
- Build/configure end-to-end MLOps and GenAI Ops frameworks: model and prompt versioning; reproducible pipelines and CI/CD; controlled deployment and rollback strategies.
- Integrate AI workflows with enterprise data platforms, orchestration tools, and cloud infrastructure.
- Model & GenAI Evaluation
- Define evaluation frameworks for ML and GenAI (accuracy/robustness/drift; LLM response quality, grounding, hallucination risk, safety checks; bias/fairness/explainability).
- Establish acceptance criteria and validation artifacts for regulated, audit-ready environments.
- Observability & Monitoring
- Implement observability to monitor ML/LLM degradation, data/embedding drift, prompt/response behavior, latency/failure modes/usage patterns.
- Enable full logging and traceability for investigations, audits, and continuous improvement.
- Responsible & Ethical AI
- Apply Responsible AI principles (human-in-the-loop controls; transparency/explainability/proper-use disclosures; privacy/access control/lineage).
- Ensure GenAI features are opt-in, governed, and aligned with AI policies and regulatory expectations.
- Collaboration & Leadership
- Partner with Data Engineering, Architecture, Security, QA, and Business teams.
- Translate business problems into well-scoped, governed AI/GenAI solutions.
- Contribute to enterprise AI standards, reference architectures, and platform roadmaps.
Requirements
- Bachelor’s or Master’s degree in Computer Science, Engineering, Data Science, or related field.
- 5+ years deploying ML systems in production.
- Strong experience with: Python; ML frameworks (PyTorch, TensorFlow, scikit-learn); LLMs/GenAI tooling; MLOps practices (pipelines/automation); cloud platforms (Azure/AWS/GCP).
- Familiarity with vector databases, embedding strategies, and RAG & graph architectures.
- Proven ability to design governed, observable, and secure AI systems.
- Experience in biotech/life sciences/healthcare or other GxP-relevant domains.
- Extensive experience with enterprise SDLC and production IT processes; full SDLC delivery of AI systems.
- Experience implementing GenAI in enterprise or regulated environments.
- Exposure to AI governance, risk assessments, or validation frameworks.