Agentic AI Engineer

Confidential Company

Agentic AI Engineering

• Design and build end-to-end agentic AI workflows using Microsoft AI Foundry and Copilot Studio, including multi-step agent pipelines capable of autonomous decision-making, tool use, and task delegation.

• Architect and deploy multi-agent systems using orchestration frameworks such as LangGraph, AutoGen, Semantic Kernel, and CrewAI, evaluating fit-for-purpose based on use case complexity and enterprise constraints.

• Implement Retrieval-Augmented Generation (RAG) pipelines incorporating vector databases, semantic search, and knowledge graph integrations to ground agent outputs in verified organisational data.

• Develop and maintain a structured prompt library, applying advanced prompting techniques including chain-of-thought, few-shot, self-reflection, and ReAct patterns to optimise agent behaviour across diverse business scenarios.

• Build and integrate Model Context Protocol (MCP) servers to enable agents to interact with internal tools, APIs, databases, and enterprise systems in a controlled and auditable manner.

• Automate end-to-end job functions across business units, including but not limited to procurement workflows, member services interactions, regulatory reporting, HR task management, document processing, and internal knowledge retrieval.

• Develop AI-powered report generation capabilities that allow agents to autonomously retrieve data, structure findings, and produce formatted outputs for business audiences without manual intervention.

• Design agent memory architectures, managing short-term conversational context and long-term persistent memory stores to enable coherent, stateful interactions across sessions.

• Build and maintain agent evaluation frameworks that measure accuracy, reliability, hallucination rates, task completion, and safety compliance, iterating on agent design based on quantitative outcomes.

• Establish guardrails, content filters, and human-in-the-loop checkpoints to ensure responsible and auditable AI agent behaviour in production environments.

• Implement LLMOps practices including model versioning, prompt version control, agent monitoring, cost tracking, and performance observability using tools such as Azure AI Foundry, LangSmith, or equivalent platforms.

• Integrate agents with enterprise systems including SharePoint, Microsoft Teams, Dynamics 365, and third-party APIs to enable seamless automation across operational workflows.

• Contribute cross-platform knowledge by evaluating and piloting emerging agentic AI tools and frameworks from across the market ensuring the organisation adopts best-in-class approaches rather than remaining constrained to a single vendor ecosystem.

Data Engineering

• Develop and maintain scalable data pipelines using Azure Data Factory, Microsoft Fabric, and Synapse Analytics to ensure agentic systems have reliable access to clean, structured, and contextually relevant data.

• Write production-grade Python and PySpark code for data transformation, orchestration, and integration tasks in support of both AI and analytical workloads.

• Implement and manage Power Automate flows to connect agentic AI outputs to downstream business processes, notifications, and approval workflows.

• Build and maintain vector stores and embedding pipelines using Azure AI Search, Cosmos DB, or equivalent technologies to support RAG and semantic retrieval use cases.

• Collaborate with the Data Engineering team to ensure data quality, lineage, and governance standards are upheld across all datasets consumed by AI agents.

• Support the integration of structured and unstructured data sources into agent-accessible knowledge layers, including document repositories, databases, and real-time event streams.

1. Required Qualifications and Experience

• Bachelor's degree or higher in Computer Science, Data Engineering, Artificial Intelligence, or a related technical discipline.

• Demonstrable experience building and deploying agentic AI systems in a production environment, including multi-agent orchestration and autonomous task execution.

• Strong proficiency in Python, including object-oriented design, API development, and integration with AI and data libraries.

• Hands-on experience with Microsoft AI Foundry, Azure OpenAI Service, and Copilot Studio.

• Practical knowledge of RAG architectures, vector databases, and semantic search implementations.

• Solid experience with Azure Data Factory, PySpark, and Microsoft Fabric or Synapse Analytics.

• Working knowledge of Power Automate for workflow automation and enterprise system integration.

• Demonstrated ability to design and maintain a structured prompt library with version-controlled, tested prompt artefacts.

• Familiarity with at least one leading agentic framework beyond the Microsoft stack, such as LangChain, LangGraph, AutoGen, CrewAI, or Semantic Kernel.