- 6+ years of professional software development or data/ML engineering experience.

- 3+ years working directly with LLMs, RAG, AI Agents or agentic AI systems development.

- MCP Server/Client Dev & Setup.

- Hands on production experience with AWS and Databricks and Bedrocker, Sagemaker in a data/AI context.

Core Responsibilities :

Agent & LLM Development :

- Build and integrate LLM based agents that can call tools/APIs, maintain state, and reason through multi step tasks. Implement retrieval augmented generation (RAG), planning, and memory mechanisms.

Data & Feature Engineering :

- Develop robust data pipelines on Databricks using Delta Lake. Prepare, transform, and curate data for LLMs and agents, including embeddings, vector indexes, and feature stores.

AWS & Platform Engineering :

- Implement scalable, secure infrastructure using AWS services (e.g., Lambda, ECS/EKS, Step Functions, S3, API Gateway, SageMaker, Bedrock Foundational Model endpoints).

- Use IaC (CloudFormation) where possible.

Databricks Engineering :

- Build and optimize notebooks, jobs, and workflows. Use Databricks ML, Model Serving, and Unity Catalog for governance, lineage, and access control. Integrate Databricks with vector databases or Lakehouse native vector search.

Agentic AI & LLM Expertise :

LLM Integration :

- Calling hosted LLM APIs (Anthropic, AWS Bedrock AI Models, etc.) or self hosted models; managing credentials, rate limits, and error handling.

RAG & Retrieval :

- Designing RAG pipelines : chunking, embedding selection, vector stores (Postgresql, S3 Vectors, or Databricks native), retrieval strategies (semantic search, hybrid search).

Agent Design :

- Multi step planning, tool use/function calling, memory and context management, managing long conversations with context windows, guardrails for safe use.

Prompt & System Design :

- System prompts, tool descriptions, response formatting, prompt optimization and experimentation for reliability and accuracy.

Evaluation & Guardrails :

- LLM evaluation techniques (human + automatic metrics), safety filters, content classification, and policy checks.

Cloud :

- AWS