Agentic AI Engineer – AWS Bedrock, LangGraph, RAG & Multi-Agent Systems

Overview

Design, develop and scale autonomous, production-grade agentic AI and generative AI systems, focused on multi-agent orchestration and advanced Retrieval-Augmented Generation (RAG) built on AWS Bedrock, working with structured and unstructured enterprise data.

What you'll do

• Design and orchestrate multi-agent workflows using planners, executors, retrievers, validators, and human-in-the-loop patterns.
• Implement advanced reasoning mechanics, including ReAct, chain-of-thought, tree-of-thought, and self-reflection loops, with appropriate safety and observability controls.
• Develop memory systems using state graphs, short-term session logs, and long-term knowledge bases to maintain state and context across multi-step workflows.
• Build autonomous task-execution engines with fine-grained tool calling, API integrations, fallback strategies, validation, and escalation controls.
• Deploy and scale foundation models through AWS Bedrock, using Knowledge Bases, Agents, Guardrails, and custom model evaluations.
• Develop secure serverless workflows using AWS Lambda and AWS Step Functions to trigger and coordinate agent actions.
• Implement LLMOps practices for prompt versioning, log streaming, latency tracking, quality measurement, and cost optimization.
• Architect end-to-end data-ingestion pipelines for embedding generation, chunking strategies, metadata tagging, and semantic indexing.
• Develop and fine-tune semantic search systems using vector databases and reranking pipelines.
• Design evaluation datasets and data pipelines to track concept drift, data drift, response quality, and automated system scores.
• Integrate unstructured enterprise repositories with LLM applications while maintaining semantic indexing accuracy, data protection, and access controls.
• Translate abstract business processes into deterministic, code-driven workflows and measurable technical outcomes.
• Collaborate with product, engineering, data, security, and operational stakeholders to define requirements and delivery plans.
• Review code and architecture, support technical decision-making, and guide team members through implementation and troubleshooting.
• Optimize AI systems for reliability, safety, scalability, latency, maintainability, sustainability, and cost.
• Maintain technical documentation covering architecture, workflows, prompts, data pipelines, evaluation methods, operating procedures, and known limitations.

What you'll need

• Expert-level Python development experience, including modular, testable, production-ready code.
• Hands-on experience with asynchronous backend APIs using FastAPI, Flask, or equivalent frameworks.
• Hands-on experience with AWS Bedrock, including model access, Knowledge Bases, Agents, Guardrails, and custom model evaluations.
• Hands-on experience with AWS EC2, Lambda, IAM, VPCs, and Step Functions.
• Deep experience with agent orchestration frameworks such as LangGraph, CrewAI, AutoGen, Semantic Kernel, or comparable frameworks.
• Experience designing multi-agent systems: planners, executors, retrievers, validators, human-in-the-loop workflows, and tool-calling loops.
• Experience designing advanced RAG systems, including ingestion, chunking, embeddings, metadata tagging, retrieval, reranking, and evaluation.
• Strong understanding of semantic search, indexing algorithms, vector databases, and reranking pipelines.
• Experience with vector technologies such as Pinecone, OpenSearch, pgvector, Chroma, or comparable.
• Experience with LLMOps: prompt versioning, log streaming, latency tracking, cost monitoring, evaluation, and operational metrics.
• Knowledge of AI safety: security guardrails, prompt-injection prevention, data alignment, access controls, and safe tool execution.
• Experience developing pipelines for structured and unstructured enterprise data, embedding generation, semantic indexing, and evaluation datasets.
• Experience across the AI application lifecycle: requirements analysis, architecture, development, testing, deployment, monitoring, and support.
• Experience developing production-level LLM and agentic AI workflows.
• Overall experience of 3–5+ years in core software development, AI engineering, backend engineering, or a related technical role.
• At least 1–2 years of focused experience developing production-level LLM and agentic AI workflows.
• Bachelor's or Master's degree in Computer Science, Information Technology, Engineering, Artificial Intelligence, or a related field; equivalent practical experience may be considered.

Nice to have

• Experience with Model Context Protocol (MCP) ecosystems or contributions to open-source agent libraries.
• Experience fine-tuning models using PEFT or LoRA techniques.
• Familiarity with model monitoring, concept drift, data drift, automated scoring, and evaluation operations.
• Experience with infrastructure as code, containerization, CI/CD, and automated deployment.
• Experience designing cost-efficient and resource-efficient AI systems.
• AWS Certified Machine Learning – Specialty or AWS Certified Solutions Architect certification.

Details

• Location: Pune - Hinjewadi (Ascendas).