Applied AI Engineer

GSK•Upper Providence, PA

About The Position

As an Applied AI Engineer, you will be embedded within cross-functional teams to deliver practical, high-impact AI/ML solutions aligned with GSK’s R&D and business priorities. You will partner closely with scientists, product teams, and domain experts to design, build, and deploy machine learning models and AI-powered tools that accelerate drug discovery, improve decision-making, and enable responsible use of AI across the enterprise. This role is hands-on and consultative in equal measure. You will evaluate use-case feasibility, prototype solutions rapidly, architect model integrations, and transfer knowledge so that partner teams can operate independently. You will also contribute to the development of reusable patterns, baseline models, and tested pipelines for common AI/ML tasks within GSK’s approved.

Requirements

Bachelor’s degree in computer science, Machine Learning, Computational Biology, Bioinformatics, Statistics, Engineering, or a related quantitative discipline; OR equivalent professional experience as a software/ML engineer.
2+ years of professional experience developing and deploying machine learning models (with a Bachelor’s); 1+ years with a Master’s or PhD.
Expertise in Python, including ML/data science libraries (PyTorch, TensorFlow, JAX, scikit-learn, pandas, numpy).
Experience with cloud platforms (GCP, AWS, or Azure) and containerization (Docker, Kubernetes).
Strong understanding of ML fundamentals: supervised/unsupervised learning, deep learning, model evaluation, feature engineering, and experiment tracking.
Experience working in healthcare, pharma, or biological domains.
Strong software engineering practices: version control (Git/GitHub), code review, testing, and documentation.

Nice To Haves

Experience in pharma, biotech, or life sciences—particularly in drug discovery, genomics, clinical data, or biological data analysis.
Hands-on experience building LLM-based applications, agentic AI systems, RAG pipelines, or multi-agent architectures (e.g., LangChain, LangGraph, AutoGen).
Experience with knowledge graph construction, causal inference, or large perturbation models.
Familiarity with single-cell RNA-seq, spatial transcriptomics, CRISPR assay data, or other high-dimensional biological datasets.
Experience with MLOps practices: CI/CD for ML, model monitoring, experiment tracking (MLflow, Weights & Biases), and reproducible research workflows.
Contributions to open-source ML/AI projects or peer-reviewed publications in applied ML.
Background or demonstrated interest in responsible AI, AI ethics, or model governance.
Experience evaluating and integrating third-party AI/ML vendor tools and platforms.

Responsibilities

Advisory & Solution Design Provide tailored guidance to business units on AI/ML use cases, feasibility, model selection, and deployment options, particularly in scientific domains without active AI/ML engineering efforts.
Co-design prototypes and proof-of-concepts (PoCs) with product and domain teams to validate ideas quickly and de-risk larger investments.
Translate complex stakeholder requirements into well-scoped technical solutions with clear success criteria and handover plans.
Model Development & Deployment Build, train, evaluate, and iterate on ML models for real-world scientific and business problems—including but not limited to NLP/LLM applications, knowledge graphs, causal inference, computer vision, and predictive modeling.
Package trained models into production-ready services (APIs, containerized deployments) using GSK’s cloud infrastructure (GCP/AWS/Azure).
Develop and maintain agentic AI systems, multi-agent architectures, and LLM-based tools where appropriate.
Share reusable patterns, baseline models, and tested pipelines for common AI/ML tasks.
Embed privacy, ethics, and regulatory considerations into every engagement from the outset.
Knowledge Transfer & Enablement Run workshops, seminars, and hands-on training sessions to increase AI literacy across the organization.
Embed within business/research units for time-limited engagements (typically 6–8 weeks) to accelerate delivery and transfer skills.
Communicate relevant issues, requests, and opportunities from business units back to AI/ML product leads.