Website:
nextamp.com
Job details:
Company Overview: NextAmp LLC is a digital modernization services company focused on transforming the insurance industry. We help insurance organizations redesign workflows and build AI-powered, automated systems that improve efficiency, accuracy, and scalability. Our solutions streamline core operational areas including claims processing, underwriting, billing, and back-office operations.
Position: Lead Agentic Developer
Location: India, Remote
Job Type: Full Time
About the role
We are looking for a hands-on Agentic Developer to design, build, and ship production-grade AI agents and intelligent workflows for our enterprise clients, including a strategic insurance engagement. You will work end-to-end across the agent lifecycle — from use case discovery and prompt design through to integration, evaluation, and production hardening — using the OpenAI and Anthropic (Claude) ecosystems alongside modern agent frameworks.
This is a builder role for an engineer who is excited about agentic AI, comfortable in ambiguity, and motivated by shipping working solutions that drive real business outcomes.
What you will do
- Design and build AI agents and multi-step workflows using frameworks such as LangChain, LangGraph, and the OpenAI Agents SDK and Anthropic Claude SDK.
- Implement Retrieval-Augmented Generation (RAG) pipelines end-to-end: ingestion, chunking, embedding, indexing, retrieval, and orchestration.
- Build tool-using agents, multi-agent orchestration patterns, and human-in-the-loop checkpoints with appropriate governance and auditability.
- Develop and harden backend services using Python (FastAPI) and integrate with frontend or downstream systems through REST APIs, Server-Sent Events, or WebSockets for real-time agent interactions.
- Own prompt engineering: design, iterate, and evaluate prompts using structured evaluation harnesses to measure quality, reduce hallucinations, and meet acceptance criteria.
- Evaluate and integrate frontier LLMs; make informed tradeoffs across quality, latency, and cost.
- Integrate agent workflows with enterprise systems, vector databases (e.g., Pinecone, Milvus, Chroma, FAISS), and where relevant, knowledge graphs such as Neo4j.
- Containerize and deploy agent services using Docker (and Kubernetes where applicable); implement monitoring and observability for agent performance, cost, and output quality.
- Partner with Project Managers, client stakeholders, and domain experts to translate workflow requirements into shippable agent solutions.
- Stay current with the agentic AI landscape and bring back ideas, patterns, and tools that improve how we build.
What we are looking for
Must have
- 3 to 6 years of experience in software engineering, with at least 1 to 2 years building AI or LLM-based applications.
- Strong Python programming skills and solid SQL fundamentals.
- Hands-on experience building agents and workflows using LangChain, LangGraph, LlamaIndex, or the native OpenAI and Anthropic SDKs.
- Practical experience implementing RAG architectures end-to-end.
- Working knowledge of prompt engineering, including iteration and evaluation techniques.
- Experience with at least one vector database (Pinecone, Milvus, Chroma, FAISS, or similar) and embedding-based retrieval.
- Experience integrating multiple LLM provider APIs (OpenAI, Anthropic, and ideally Google GenAI) and benchmarking across them.
- API development with FastAPI or equivalent; comfortable with REST design and asynchronous patterns.
- Experience with Docker and scalable runtime patterns.
- Familiarity with at least one cloud provider's AI/ML services (Azure, AWS, or GCP).
Nice to have
- Experience with real-time and streaming patterns (Redis pub/sub, Server-Sent Events, WebSockets) for agent task orchestration.
- Exposure to knowledge graphs (e.g., Neo4j) for structured reasoning.
- Experience with model evaluation frameworks, observability tooling for LLMs, or agent tracing tools (e.g., LangSmith, Langfuse).
- Insurance, financial services, or other regulated-industry exposure.
- Familiarity with responsible AI practices, guardrails, and enterprise governance patterns.
- Contributions to open-source agentic AI projects or a public portfolio of agent work.\
Working Hours : The ideal candidate will be working from 1:30pm IST - 10:30pm IST
Click on Apply to know more.