Job Summary
10+ years (min. 2 years in GenAI / LLM systems)
We are seeking a Senior AI/ML Engineer to design, develop, and deploy production-grade AI/ML solutions within GSK’s Digital & Tech organization. This role focuses on building Generative AI applications, multi-agent systems, and advanced retrieval pipelines that drive measurable business impact. You will work at the intersection of cutting-edge AI research and enterprise software engineering, collaborating with data scientists, platform engineers, and domain experts across R&D, supply chain, and commercial functions.
Key Responsibilities
Generative AI & LLM Development
- Design, develop, and deploy Generative AI applications using LLMs (GPT-4, Claude, Gemini, open-source models) for enterprise use cases
- Build and orchestrate multi-agent systems using frameworks like LangGraph, LangChain, CrewAI, or AutoGen with function calling and tool use
- Implement Retrieval-Augmented Generation (RAG), Graph RAG, and hybrid retrieval pipelines using vector databases (Pinecone, Weaviate, Chroma, pgvector)
- Apply prompt engineering, chain-of-thought reasoning, and context engineering techniques to optimize model outputs
- Fine-tune LLMs and embedding models for domain-specific tasks using LoRA, QLoRA, or full fine-tuning approaches
- Implement guardrails, content filtering, and safety mechanisms for responsible AI deployment
ML Engineering & MLOps
- Build end-to-end ML pipelines – data ingestion, feature engineering, model training, evaluation, and deployment
- Implement LLMOps practices: model versioning, A/B testing, prompt management, evaluation frameworks (LLM-as-judge, RAGAS, custom metrics)
- Deploy and manage LLM inference using frameworks such as vLLM, TensorRT-LLM, or DeepSpeed for latency and cost optimization
- Monitor model performance, detect drift, and implement continuous improvement loops
- Build observability for AI systems using LangSmith, Langfuse, or custom tracing solutions
Architecture & Cloud
- Architect scalable AI solutions on AWS (Bedrock, SageMaker, Lambda) or Azure (OpenAI Service, ML Studio)
- Containerize AI applications with Docker and deploy via Kubernetes, ECS, or serverless patterns
- Design event-driven and API-first architectures for AI service integration with enterprise systems
- Implement CI/CD pipelines for ML models and AI applications
Collaboration & Leadership
- Collaborate with data scientists, domain experts, and product owners to translate business problems into AI solutions
- Conduct code reviews, architectural design reviews, and contribute to engineering standards
- Mentor junior AI/ML engineers; lead technical knowledge-sharing sessions
- Evaluate and recommend emerging AI technologies, frameworks, and approaches
- Present AI solutions and results to technical and non-technical stakeholders
Skill Requirements
|
Skill Area |
Required Proficiency / Technologies |
|
GenAI & LLMs |
LangChain, LangGraph, OpenAI API, Claude API, Hugging Face Transformers, prompt engineering, multi-agent orchestration |
|
ML Frameworks |
PyTorch, TensorFlow, Scikit-learn, XGBoost; fine-tuning (LoRA / QLoRA / PEFT) |
|
NLP & Retrieval |
RAG, Graph RAG, hybrid search, vector DBs (Pinecone, Weaviate, Chroma), embedding models, NER, text classification |
|
LLMOps & Eval |
LangSmith, Langfuse, RAGAS, LLM-as-judge, model versioning, A/B testing, prompt management |
|
Inference |
vLLM, TensorRT-LLM, DeepSpeed, ONNX Runtime, quantization techniques |
|
Cloud & Infra |
AWS (Bedrock, SageMaker, Lambda, S3) or Azure; Docker, Kubernetes, Terraform |
|
Programming |
Python (primary), FastAPI, SQL, Git, Bash; familiarity with TypeScript / JavaScript a plus |
|
Data & Tools |
PostgreSQL, Neo4j, MongoDB, Redis, Apache Kafka, Databricks, Jupyter, MLflow |