Agentic AI Engineer

Job Description: Agentic AI Engineer
Experience: 3–6 Years
Role Overview
The Agentic AI Engineer will be responsible for designing, developing, and deploying AI-driven solutions from initial prototypes to production environments. The role involves working with Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), AI agents, tool integration, orchestration frameworks, cloud-based AI services, and modern software engineering practices.
Key Responsibilities and Requirements
A minimum of 4 years of experience in AI engineering, software development, data engineering, machine learning, cloud engineering, or related technical fields.
Hands-on experience developing Generative AI applications, AI agents, RAG solutions, enterprise search systems, copilots, and LLM-based workflow automation.
Strong proficiency in Python, including API development, backend services, automation scripts, and reusable AI components.
Solid understanding of LLM concepts, including prompt and context engineering, model selection, temperature and top-p parameters, context windows, embeddings, token consumption, response latency, and cost optimization.
Practical experience building RAG pipelines using vector databases, embedding models, document processing, retrieval techniques, metadata filtering, grounding, and citation-based responses.
Experience implementing multi-agent architectures, including supervisor agents, planner-executor workflows, routing agents, tool-using agents, evaluator agents, and human-in-the-loop processes.
Ability to integrate AI agents with databases, APIs, enterprise applications, documents, and external services through tool calling.
Understanding of agent memory mechanisms, including session-based memory, long-term memory, vector storage, user context, conversation history, and controlled memory retention.
Experience with AI evaluation techniques covering response accuracy, grounding, hallucination detection, retrieval performance, regression testing, adversarial testing, and user feedback.
Knowledge of AI governance and responsible AI practices, including model approval processes, selection criteria, security controls, evaluation documentation, auditability, and lifecycle management.
Experience implementing agent guardrails such as policy enforcement, restricted tool access, approval workflows, fallback mechanisms, escalation procedures, human oversight, and emergency shutdown controls.
Familiarity with observability and tracing for AI systems, including agent execution logs, tool-call tracking, prompt and response metadata, token usage, latency monitoring, error analysis, and production troubleshooting.
Hands-on experience with agent development frameworks such as LangChain, LangGraph, LlamaIndex, Semantic Kernel, CrewAI, AutoGen, or equivalent technologies.
Experience with cloud-based AI platforms, including AWS Bedrock Agents, AWS SageMaker, Azure OpenAI, Azure AI Agent Service, Azure AI Foundry, or similar services.
Understanding of enterprise data architecture, including structured and unstructured data, semantic layers, data catalogues, metadata management, data quality, and access governance.
Strong knowledge of REST APIs, microservices, authentication, secrets management, logging, and cloud-native application development.
Understanding of AI security principles, including role-based access control, data privacy, prompt injection prevention, hallucination mitigation, content filtering, audit trails, and secure agent execution.
Ability to collaborate with business stakeholders to identify requirements and translate business needs into practical AI agent solutions.
Strong communication and collaboration skills, with the ability to work effectively with architects, data engineers, platform teams, product owners, and business subject-matter experts.
Nice-to-Have Skills
Familiarity with AI observability and tracing tools such as LangSmith, Arize Phoenix, OpenTelemetry, MLflow Tracing, Databricks MLflow, or equivalent cloud monitoring solutions.
Experience designing human-supervised AI workflows involving approval mechanisms, exception handling, escalation procedures, feedback collection, and controlled autonomy.
Knowledge of model risk management, responsible AI governance, prompt management, model catalogues, evaluation reporting, and audit-ready documentation.
Experience building reusable tool registries, plugin architectures, MCP-based integrations, OpenAPI integrations, and schema-driven tool invocation.
Exposure to advanced multi-agent coordination patterns involving supervisor, planner-executor, critic/evaluator, router, and specialized task agents.
Experience defining structured tool interfaces using OpenAPI specifications, JSON Schema, function-calling definitions, structured outputs, API contracts, and input validation mechanisms.

Work arrangement
No