ML Engineer – Agentic AI

About the job

Responsibilities • Design and implement multi-agent systems using LangChain, LangGraph, CrewAI, AutoGen or similar frameworks. • Build A2A (agent-to-agent) orchestration and implement MCP (multi-context protocol) for context reuse and collaboration. • Fine-tune foundation models (LoRA, RLHF, DPO) and integrate retrieval-augmented generation (RAG) pipelines. • Develop APIs and backend services (FastAPI / Flask / Node.js) and integrate with enterprise systems (HRIS, CRM, ERP). • Build secure data pipelines (ETL, embeddings, vector DB ingestion) and ensure governance, bias mitigation, and lineage. • Containerize, deploy, and monitor agents (Docker, Kubernetes, CI / CD pipelines) with observability and drift detection. • Collaborate with QA, red teaming, and product leads to benchmark accuracy, safety, and compliance. Skills Required • Strong in Python, PyTorch / TensorFlow, Hugging Face, and fine-tuning techniques (LoRA, RLHF). • Hands-on with LangChain / LangGraph, MCP, A2A orchestration, and multi-agent architectures. • Proficient in ETL pipelines, SQL / NoSQL, vector DBs (Pinecone, Weaviate, FAISS, Milvus). • Backend expertise in FastAPI / Flask / Node.js, REST / GraphQL APIs, and microservices. • DevOps / MLOps experience : Docker, Kubernetes, CI / CD (Jenkins / GitHub Actions), monitoring / observability. • Cloud experience with AWS / GCP / Azure AI services and secure integration practices. • Bonus : familiarity with enterprise HR / ATS systems (Workday, SAP, Paychex) and workflow automation.

Requirements

• Python
• PyTorch
• TensorFlow
• Hugging Face
• ETL pipelines
• FastAPI
• Flask
• Node.js
• DevOps
• MLOps

Preferred Technologies

• Python
• PyTorch
• TensorFlow
• Hugging Face
• ETL pipelines
• FastAPI
• Flask
• Node.js
• DevOps
• MLOps