Lead Agentic AI Engineer
accenture
Job Description
- Design, build, and deploy production-grade AI, Generative AI, and Agentic AI applications on Google Cloud, primarily using Vertex AI and Gemini models.
- Develop intelligent AI applications and agents capable of reasoning, retrieval, tool use, workflow orchestration, structured output generation, task automation, and enterprise system integration.
- Build scalable AI application architectures integrating Vertex AI with GCP services such as BigQuery, Cloud Storage, Cloud Run, GKE, Pub/Sub, API management, databases, and enterprise applications.
- Apply strong software engineering principles to develop secure APIs, microservices, AI services, data pipelines, agent tools, and reusable AI components suitable for enterprise production environments.
- Design, develop, test, and deploy Generative AI, LLM, Machine Learning, and Agentic AI solutions using Google Cloud Platform and Vertex AI.
- Build applications using Vertex AI, Gemini models, Vertex AI APIs, embeddings, model endpoints, prompt management, grounding, function/tool calling, and other GCP AI capabilities.
- Develop AI agents capable of planning, reasoning, tool calling, information retrieval, workflow execution, memory management, and multi-step task automation.
- Design and implement Retrieval-Augmented Generation (RAG) solutions using Vertex AI, embeddings, vector search, enterprise documents, structured data, semantic search, and appropriate retrieval strategies.
- Build integrations between AI applications and GCP services such as BigQuery, Cloud Storage, Cloud Run, Cloud Functions, GKE, Pub/Sub, Secret Manager, and other cloud-native services.
- Develop backend APIs, microservices, connectors, integration services, and reusable tools that allow AI applications and agents to interact securely with enterprise systems, databases, APIs, and external services.
- Implement Model Context Protocol (MCP) clients or servers where applicable to provide standardized and secure access to tools, APIs, enterprise applications, and data sources.
- Work with Agent2Agent (A2A) patterns or protocols for agent discovery, task delegation, inter-agent communication, and multi-agent collaboration where required.
- Work with AI/LLM orchestration frameworks such as Google Agent Development Kit (ADK), LangChain, LangGraph, LlamaIndex, Semantic Kernel, AutoGen, CrewAI, or equivalent technologies.
- Evaluate and improve AI application quality across accuracy, groundedness, hallucination reduction, prompt quality, retrieval quality, latency, reliability, scalability, security, and cost efficiency.
- Implement logging, monitoring, tracing, observability, evaluation, guardrails, and production support mechanisms for AI applications and agentic workflows.
- Collaborate with architects, product owners, data engineers, backend developers, ML engineers, security teams, and DevOps teams to deliver enterprise-grade AI solutions.
- Follow software engineering best practices including Git-based development, automated testing, code reviews, CI/CD, infrastructure automation, documentation, security, and production release management.