Agentic AI Architecture & Development

Role

Lead architecture and development of production-grade agentic AI systems and backend services. Implement multi-agent orchestration, RAG pipelines, LLM integrations, evaluation and observability, and operational patterns required to run agentic workloads at enterprise scale.

Key Responsibilities

• Design and build multi-agent systems using LangGraph as the primary orchestration framework; apply agent orchestration patterns (planning, tool use, persistent state, memory, reflection, multi-agent coordination).
• Develop and optimize RAG (Retrieval-Augmented Generation) pipelines: document processing, chunking strategies, embedding workflows, and vector database integration.
• Integrate and manage LLM/SLM services (OpenAI, Azure OpenAI, Anthropic, open-source models) with model selection, prompt engineering, and cost optimization.
• Design prompt engineering strategies (chain-of-thought, few-shot, structured outputs) and implement guardrails, safety mechanisms, and content filtering.
• Build scalable Python backend services (FastAPI) to serve AI agent workflows; implement caching, rate limiting, persistent agent state, and conversation memory.
• Develop event-driven microservices, real-time streaming, distributed task processing (Celery), and event-driven autoscaling (KEDA) for production workloads.
• Build agent evaluation, testing, and observability frameworks to ensure reliability and performance in production.
• Lead proofs-of-concept, drive successful experiments to production, mentor engineers, and contribute to architecture decision records and technical documentation.

Requirements

• 6+ years of software engineering experience with significant hands-on AI/ML work in enterprise environments.
• Strong proficiency in Python and experience building production AI applications and APIs.
• Demonstrated experience with LangGraph or similar agentic AI frameworks (LangChain, CrewAI, AutoGen) in production systems.
• Hands-on experience with LLM API integration (OpenAI, Azure OpenAI, Anthropic) and prompt engineering.
• Experience designing and implementing RAG systems including embedding models and vector database integration.
• Experience with Python web frameworks (FastAPI), distributed task processing (Celery), event-driven microservices, and real-time streaming.
• Proficiency with Git, CI/CD practices, and cloud platforms (preferably Azure).
• Strong communication skills and familiarity with Agile methodologies.

Preferred Qualifications

• Experience with vector databases such as Qdrant, Pinecone, Weaviate, ChromaDB.
• Experience with Databricks Genie or similar natural language-to-data query platforms.
• Familiarity with AWS Bedrock AgentCore for managed agent runtimes and multi-cloud deployments.
• Knowledge of model fine-tuning, quantization, serving optimization, multi-tenant patterns, containerization (Docker, Kubernetes), and KEDA.
• Understanding of AI safety, responsible AI principles, and enterprise governance.

Tech & Tools

• Primary frameworks and platforms: LangGraph, LangChain, CrewAI, AutoGen
• LLM providers and services: OpenAI (GPT-5.X referenced), Azure OpenAI, Anthropic
• Backend & orchestration: Python, FastAPI, Celery, KEDA
• Vector DBs & data: Qdrant, Pinecone, Weaviate, ChromaDB, Databricks Genie
• Dev tools: Cursor AI, GitHub Copilot, Git, CI/CD
• Cloud & runtimes: Azure (preferred), AWS Bedrock AgentCore

Employment & Location

• Employment type: Full-time
• Location: Bangalore (Hybrid Role)
• Contact: Share resume to Vishal Kumar at Vishal@akaasa.com