Software Engineer III AI Tools

Framework Ventures; New Delhi; On-site; INR 5,500,000 - 7,300,000; Full time

Job summary

Zinnia is seeking an experienced GenAI Engineer to design, build, and operate production LLM systems. You will own backend services for GenAI features in Python/FastAPI or TS/Node, from design through production.

Build end-to-end RAG pipelines with embeddings, vector indexes, retrieval, reranking, grounding, and response synthesis. You will also work on MCP integrations, LLM evaluation and guardrails, and stand up LLMOps/MLOps, including CI/CD, dataset/version management, feature stores,

Qualifications

• 5-9 years of experience in backend/GenAI engineering.
• Strong backend skills with Python or TS/Node.js.
• Hands-on with LLM prompts, RAG pipelines, and evaluation.
• Experience with vector stores and embeddings pipelines.
• Familiarity with MLOps, CI/CD, and model/versioning.

Responsibilities

• Own backend services for GenAI features from design to production.
• Build end-to-end RAG pipelines: chunking, embeddings, indexes, retrieval, ranking.
• Integrate MCPs for internal apps/apis; handle auth and rate limits.
• Implement LLM evaluation, guardrails, A/B tests, and safety checks.
• Stand up LLMOps/MLOps: CI/CD, dataset/versioning, observability.

Skills

Python

TypeScript/Node

FastAPI

LLM prompting

Vector search

MLOps

Kubernetes

Docker

CI/CD

RBAC

Tools

Docker

Kubernetes

AWS

GCP

OpenAI SDKs

Job description

WHO WE ARE:

Zinnia is the leading technology platform for accelerating life and annuity growth. With innovative enterprise solutions and data insights, Zinnia simplifies the experience of buying, selling, and administering insurance products. All of which enables more people to protect their financial futures. Our success is driven by a commitment to three core values: be bold, team up, deliver value – and that we do. Zinnia has over $180 billion in assets under administration, serves 100+ carrier clients, 2500 distributors and partners, and over 2 million policyholders.

WHO YOU ARE:

Generative AI Engineer will be responsible to design, build, and operate production LLM systems—RAG services, eval/guardrail pipelines, and API integrations—used across the enterprise.

WHAT YOU'LL DO:

Own backend services for GenAI features (Python/FastAPI or TypeScript/Node), from design to production.

Build RAG pipelines end-to-end: chunking, embeddings, vector indexes, retrieval, reranking, grounding, and response synthesis.

Build/extend MCP integrations for major internal applications and APIs.

Implement LLM evaluation & guardrails: prompt/unit evals, Ragas, Langfuse, LangSmith, A/B tests, hallucination & safety checks, feedback loops.

Stand up LLMOps/MLOps: CI/CD for models & prompts, dataset/version management, feature stores, tracing, monitoring, and cost controls.

Create SDKs/REST/GraphQL integrations with internal systems; handle auth (OAuth/JWT), rate limits, retries, and backoff.

Drive observability (metrics, tracing, logs), SLOs, and incident response; partner with SRE for reliability.

Collaborate with product and security on requirements, data governance, and privacy-by-design.

Mentor peers; raise code quality through reviews, design docs, and reusable components.

WHAT YOU’LL NEED:

Total 5-9 years of experience is required.

Backend strength: Python (FastAPI, Pydantic, async) or TypeScript/Node (Express/Fastify/Next API routes); testing (pytest/jest), Git/PR hygiene, CI/CD.

GenAI hands-on: LLM prompting/tool use, context management, streaming, and at least one RAG system taken to production (or a rigorous POC with metrics).

Vector search: FAISS/pgvector/Pinecone/Weaviate/Milvus/Qdrant; embeddings pipelines; caching & cold-start strategies.

Ops for AI: Practical MLOps/LLMOps/AIOps—model/prompt versioning, data contracts, evaluation harnesses, cost/perf monitoring, rollbacks.

Cloud & infra: Docker, Kubernetes (or serverless), one major cloud (AWS/Azure/GCP), secrets management, IaC basics.

Security & governance: PII handling, RBAC, audit trails, and least-privilege design.

Nice to have

• LangChain/LlamaIndex/LangGraph, Agentic AI, OpenAI/Azure, Gemini, OpenAI/Bedrock/Vertex SDKs; MCP (Model Context Protocol) tools/adapters.
• Retrieval enhancements: hybrid search, re-rankers, prompt caching, toolformer/function calling.
• Data systems: PostgreSQL, Redis, Kafka/SQS; batch & streaming pipelines.
• Observability: OpenTelemetry, Prometheus, Grafana; latency/cost optimization.
• Basic frontend for internal tools (React/Next.js) to wire quick admin consoles.

WHAT’S IN IT FOR YOU?

At Zinnia, you collaborate with smart, creative professionals who are dedicated to delivering cutting‑edge technologies, deeper data insights, and enhanced services to transform how insurance is done.

We are an Equal Opportunity employer committed to a diverse workforce. We do not discriminate based on race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability.