Job Summary The AI Engineer is responsible for building, deploying, and operating production-grade AI and Agentic AI solutions across the enterprise. This is a hands-on engineering role focused on implementing LLM-powered applications, orchestration and multi-agent workflows, and secure API-driven integrations using Azure AI Foundry and Azure-native services (compute, storage, messaging, data, security, and observability).

The role works across the full AI lifecycle from system design and development to production operations ensuring solutions are secure, scalable, observable, and governed. The AI Engineer partners closely with Engineering Leads, architects, data science teams, and platform/security stakeholders to translate AI use cases into reliable, enterpriseready systems, not isolated proofs of concept.

A key expectation of the role is to embed evaluation, quality monitoring, and runtime security into AI systems, including the use of LLM-as-a-Judge patterns and alignment to agent lifecycle governance and runtime protection controls (e.g., Agent 365aligned environments).

Essential Functions of the Job:

- Build Agentic AI Solutions using Azure AI Foundry (Core): Build and evolve AI applications and agents leveraging Azure AI Foundry-aligned capabilities used in the enterprise toolchain
- Develop Orchestration & MultiAgent Systems (Core): Implement orchestration layers to coordinate tools/agents across multi-step tasks, including multi-agent workflows for specialized sub-tasks and coordinated execution.
- FullStack Engineering + API Development (Core): Build end-to-end AI experiences (UI where applicable), backend services, and integration layers. Design and implement RESTful APIs and microservices that expose AI/agent capabilities securely and reliably.
- Serverless & Asynchronous Processing with Azure Functions (Core): Build services using Azure Functions, including Durable Functions (or equivalent)



for long-running and stateful orchestration patterns.
- Messaging / Queues for Workflow Reliability (Core): Use queue/event-driven patterns (e.g., messaging, pub/sub) to decouple services and improve reliability of multi-step AI pipelines and orchestration flows.
- Data Engineering Foundations: ADLS + Retrieval (Core): Work with ADLS-style data lake patterns for ingestion, storage, and processing to support AI workloads and grounding.
- Use Azure Cognitive Services / Azure AI Services Where Appropriate (Core: Leverage Azure Cognitive Services / Azure AI Services capabilities as part of end-to-end AI solutions.
- Vector Databases & Knowledge Stores (Core): Implement vector retrieval using enterprise options such as Azure AI Search and/or other vector DB patterns (e.g., Cosmos DB / PostgreSQL pgvector / Redis) based on operational needs.
- Continuous Evaluation using LLM-as-a-Judge (Core): Implement evaluation pipelines for LLM/agent outputs, including LLM-as-a-Judge patterns and structured scoring/assessment approaches where appropriate.
- Agent Lifecycle Governance with Agent 365 (A365) Awareness (Core): Build solutions that align with enterprise lifecycle management and governance patterns such as:
- Agent registry / inventory expectations
- Access controls and telemetry/observability requirements
- Monitoring and operational controls at agent scale

- Runtime Security / Runtime Protection (Core): Implement and support runtime protection expectations for agentic solutions, and participate in controls aligned to:





- Runtime protection
- Access controls (e.g., Entra ID patterns)
- Threat detection and monitoring expectations

Analytical/Decision Making Responsibilities:
- This role is critical to ensuring the enterprises AI ambition translates into real, reliable, and scalable systems not just innovation theater. You will define how AI is built, shipped, and operated across the organization.

Knowledge and Skills Requirements: Core Software Engineering (Required)
- Solid hands-on development in Python / C# / TypeScript/JavaScript (or similar).
- Experience building API-driven services and integrating distributed systems.
- Strong understanding of non-functional requirements: reliability, availability, scalability, performance, and cost.

Azure & Platform Engineering
- Handson experience with Azure AI Foundry for delivering enterprise GenAI solutions in an enterprise context.
- Experience building serverless and asynchronous workloads using Azure Functions (including Durable Functions or equivalent).
- Experience using queues and eventdriven messaging for decoupled, reliable workflows.
- Experience working with ADLS for data ingestion, storage, and processing.
- Experience using Azure Cognitive Services / Azure AI Services as part of AI solutions.

Azure Functions + Queues/Eventing (Required)
- Experience with Azure Functions (including Durable Functions or equivalent orchestration patterns).
- Experience implementing asynchronous patterns with queues/eventing for scale and reliability.

Vector Databases (Required)
- Hands-on experience with vector databases / vector search, including enterprise deployment patterns

LLM-as-a-Judge Evaluation (Required)
- Experience implementing evaluation approaches that include LLM-as-a-Judge (or equivalent automated evaluation patterns) for quality monitoring and continuous improvement.

📌 AI Engineer - Agentic AI & Azure AI Foundry (Kochi)
🏢 EY
📍 Kochi