Role & responsibilities

- Client Embedding & Workflow Discovery: Partner directly with business stakeholders to break down manual enterprise processes, map complex decision-making steps, and translate them into robust agentic workflows.
- Agentic AI Engineering: Design, build, and deploy custom multi-agent systems, function-calling pipelines, RAG implementations, and reusable agent skills.
- End-to-End System Ownership: Develop production-ready backend services and integration layers to connect AI agents with core enterprise platforms (e.g., SharePoint, Jira, SAP, email).
- Cloud & Infrastructure Management: Package, host, and scale agent workloads using contemporary cloud architectures (AWS AgentCore, serverless, containerized environments) with automated CI/CD.
- Safety & Observability: Implement guardrails, human-in-the-loop checkpoints, and agent evaluation loops to ensure reliability, mitigate hallucinations, and ensure enterprise compliance.
- Frameworks:



Hands-on experience with Strands (primary framework), along with

familiarity in LangGraph, CrewAI, or AutoGen.

- Agent Core Concepts: Deep understanding of tool use, function calling, streaming, context management, system prompts, and chain-of-thought methodologies.
- MCP & Agent Protocols: Proven ability to build/consume Model Context Protocol (MCP) servers, alongside multi-agent coordination (A2A), task delegation, and result aggregation.
- RAG Pipelines: Hands-on build experience with vector databases (e.g., Pinecone, pgvector, OpenSearch), chunking strategies, embeddings, and retrieval optimization.
- Skill Building & Guardrails: Design reusable skills for agent consumption; implement observability, hallucination mitigation, and human-in-the-loop checkpoints.

Preferred candidate profile

Perks and benefits

📌 Forward Deployed Engineer (Pune)
🏢 Ciklum
📍 Pune