Python with Agentic AI

JOB DESCRIPTION

Python Developer – Agentic AI Platform Calpion Inc. | Healthcare RCM Automation | Confidential

Location: Bangalore, India
Experience: 2–8 Years .

About the Role

We are building an enterprise Agentic AI platform to automate healthcare Revenue Cycle Management (RCM) workflows across Artemis and Olympus. The platform combines deterministic automation with LLM-powered agents to handle claims processing, denial management, payer portal interactions, appeal generation, EDI transactions, and document processing.

This is a production engineering role focused on building intelligent, self-healing agents—not chatbots or copilots.

Key Responsibilities

• Design and develop multi-agent workflows using LangGraph and other agent frameworks.

• Build reusable MCP servers and tools for healthcare APIs, clearinghouses, and payer integrations.

• Develop Playwright-based browser automation with LLM-powered self-healing capabilities.

• Integrate AWS Step Functions for stateful workflows and human-in-the-loop approvals.

• Implement reliable, scalable automation for healthcare claims, denials, and payer intelligence.

Must-Have Skills

Python: 4+ years of production Python, asynchronous programming (asyncio, aiohttp), Pydantic, and web automation using Playwright (preferred), Selenium, or Pyppeteer.

AWS: Hands-on experience with at least four services, including ECS Fargate, Lambda, Step Functions, SQS, SNS, EventBridge, S3, RDS PostgreSQL, Secrets Manager, or Bedrock. Knowledge of IAM, VPC, and CDK/CloudFormation.

LLM / Agentic AI: Practical experience with Claude, OpenAI, or equivalent LLM APIs; LangGraph, LangChain, CrewAI, or AutoGen; prompt engineering; RAG; and structured outputs.

Software Engineering: REST APIs, PostgreSQL, Git, CI/CD, error handling, retry mechanisms, structured logging, and CloudWatch.

Good-to-Have Skills

• Healthcare RCM knowledge: EDI 835/837, CARC/RARC, CPT/HCPCS, and payer workflows.

• MCP server development and Amazon Bedrock AgentCore.

• LangGraph state-machine design and multi-agent orchestration.

• Document AI, PDF extraction, OCR, and table extraction.

• HIPAA compliance and PHI security awareness.

• Playwright MCP, voice integrations (Twilio, Deepgram, ElevenLabs), and RPA tools such as UiPath or Automation Anywhere.

What We Are NOT Looking For

• Pure data scientists or ML researchers focused on model training.

• Chatbot or copilot developers without production automation experience.

• Candidates who know frameworks but lack sound engineering judgment in choosing between LLMs, deterministic scripts, and agent-based workflows.