GenAI Developer/Engineer
Experience: Total Experience: 10-15 years, AI-ML and GenAI Experience: 4–6 years
Strong fundamentals and realworld production exposure. Responsible for building, deploying, and operating GenAI solutions using modern agentic architectures.
Core Responsibilities
- Design, develop, and deploy GenAI and ML solutions using Azure OpenAI, Azure ML, Azure AI Studio, and related services
- Build LLMpowered applications using: Prompt engineering, Tool/function calling, RetrievalAugmented Generation (RAG) and Agentic AI design patterns
- Develop AgenttoAgent (A2A) workflows for task decomposition, orchestration, and collaboration
- Implement Model Context Protocol (MCP) to standardize tool access, memory, and context management across agents
- Apply MLOps and LLMOps practices:
- CI/CD for models and prompts and Versioning, monitoring, evaluation
- Failure detection and mitigation (hallucinations, drift, prompt regressions)
- Implement LLM security and guardrails:
- Prompt injection prevention,
- Data leakage protection and Output moderation and policy enforcement
- Finetune and adapt models for economy - token economics, latency, and cost optimization
- Build and expose AI capabilities via secure APIs and microservices
- Collaborate with architects, data engineers, and platform teams
- Actively research and document emerging GenAI techniques and tools
Mentor junior engineers and contribute to internal knowledge assets
Technology Stack
- Programming: Python (PEP8 / Google Python standards)
- LLMs: Azure OpenAI, OpenAI, opensource and local LLMs
- AI Platforms: Azure ML, Azure Databricks, Azure AI Services
- Agent Frameworks: LangChain, LangGraph, Semantic Kernel (or equivalent)
- Protocols & Patterns: MCP, A2A, RAG variants
- DevOps: Docker, REST APIs, CI/CD pipelines
- Datastores: SQLbased RDBMS, Vector databases
- Graph databases: Neo4j, GraphDB
Soft Skills
- Strong ownership mindset and delivery focus
- Ability to work in agile, fastchanging environments
Effective collaboration across engineering and architecture teams
