AI / ML Engineer

Build and evolve a production-grade, multi-agent conversational search platform built on large language models and enterprise semantic search.

Deployment of advanced AI/ML models and data solutions that drive strategic decision-making and operational efficiency across GRA services.

Collaborate with cross-functional teams to deliver scalable, secure, and compliant products.

Responsibilities:

• Design and implement multi-agent architectures using LLM orchestration frameworks (e. g., LangChain, LangGraph, Strands) with Amazon Bedrock or equivalent managed LLM services.
• Build and tune conversational agents with memory management strategies (e. g., sliding window and summarisation-based context compression).
• Integrate enterprise search indices (Amazon Kendra or similar) with generative AI layers for RAG pipelines.
• Develop prompt engineering patterns, guardrails, and content filtering mechanisms for production-grade AI safety.
• Implement document analysis pipelines capable of processing PDFs, Word documents, and scanned images (OCR, format conversion).
• Build ranking and relevance scoring systems using LLMs in parallel processing architectures.
• Optimise Bedrock / LLM API usage for latency, cost, and token efficiency.
• Collaborate with data engineers on embedding data pipelines using AWS and modern lakehouse architecture.
• Collaborate closely with backend and frontend engineering teams to integrate agentic AI systems into user-facing applications.
• Define evaluation frameworks for agent response quality, relevance, and safety.
• Automation & DevOps: Implement CI/CD pipelines, automate workflows, and ensure model monitoring and retraining.

Requirements:

• LLM / AI Frameworks: Amazon Bedrock, OpenAI, and Anthropic Claude; agent frameworks (Strands, LangChain, AutoGen, or similar).
• RAG Architecture: Vector search, hybrid search, Kendra / OpenSearch integration.
• Deep understanding of machine learning models and experience with deploying models to unified model hosting platforms to solve business problems.
• Experience with GenAI and NLP applications. Strong proficiency in Python.
• AWS AI/ML Services: Bedrock, Kendra, Textract, and Comprehend.
• Prompt Engineering: System prompts, few-shot, chain-of-thought.
• Conversational Memory: Context window management and summarisation strategies.
• API Design: RESTful APIs and API Gateway.
• Containerisation: Docker, AWS container-based platforms, EKS.

Nice to Have:

• Experience with multi-modal document processing (PDF, image, Word).
• Knowledge of content moderation and AI guardrails.
• Familiarity with CloudFormation / SAM for infrastructure deployment.
• Experience in regulated industries - healthcare, pharma.