Senior AI Engineer (Generative & Agentic AI)

Role Overview

We are looking for a hands-on Senior AI Engineer to build and deploy cutting-edge Generative AI applications. You will be responsible for the code-level implementation of autonomous agents, RAG pipelines, and custom model fine-tuning. Working closely with architects and data teams, you will turn conceptual designs into robust, production-grade software using Amazon Bedrock and modern Python frameworks.

Key Responsibilities

· Agent Development: Build and deploy sophisticated AI agents using Python, LangChain, and LangGraph. Implement tool calling (function calling), memory persistence, and error-handling logic for robust autonomous execution.

· RAG Implementation: Develop high-performance retrieval pipelines. Optimise chunking strategies, embedding generation (using Amazon Titan or similar), and vector search retrieval in Amazon OpenSearch.

· Model Fine-Tuning: Execute fine-tuning jobs on Amazon SageMaker, preparing training datasets and optimising hyperparameters to improve model performance on domain-specific tasks.

· Performance Engineering: Optimise latency and throughput of AI applications. Implement caching strategies and use SageMaker Endpoints for efficient inference scaling.

· Code Quality & DevOps: Write clean, testable Python code. Build CI/CD pipelines for AI models and manage prompt versioning and evaluation metrics.

Technical Skills & Requirements

· Programming: Expert proficiency in Python (Pydantic, AsyncIO, API development with FastAPI/Flask).

· AWS AI Stack: Hands-on experience with Amazon Bedrock APIs, SageMaker (Training Jobs, Inference Endpoints), and Vector Databases.

· Frameworks: extensive experience with LangChain, LlamaIndex, or LangGraph for building agentic workflows.

· LLM Techniques: Practical experience with Prompt Engineering (Chain-of-Thought, ReAct), RAG optimisation, and model evaluation techniques.

Advantageous Skills

· Middleware Integration: Practical experience building recipes or connectors in Workato to trigger AI workflows from external business events is a strong plus.

· Data & Traditional ML: Experience with Databricks for data engineering (PySpark) or building classical ML models (Scikit-Learn, XGBoost) is beneficial for handling complex data pre-processing and hybrid AI use cases.