AI Engineer

Position : AI Engineer (Platform & Product)

Location : Onsite, Delhi

Experience : 2 to 5 years

Package : 8 to 12 LPA

About the Role

We're looking for an AI Engineer who can build AI-powered product features and internal developer workflows on top of our existing MERN-stack platform. This role is for someone who doesn't just integrate an API, but can design reliable AI systems end-to-end: from prompt + evaluation, to RAG, to agents/workflows, to deployment, monitoring, and iteration.

We already use modern AI coding tools (Cursor, Claude Code, CodeRabbit, etc.) and have internal workflows to boost engineering throughput. You'll help us take AI-native engineering to the next level—both in the product and in the way our engineering team builds.

AI Engineer (LLM + Platform) — Build AI That Ships in 30 Days (CarePro)

CarePro is building intelligent software solutions to help healthcare providers run smoother operations and deliver better patient care.

Designed specifically for hospitals, CarePro enables improved patient conversion, increased revenue, and enhanced service delivery through a unified One Patient, One View approach. Our platform provides a comprehensive 360° view of the patient journey — from lead management to admission, treatment, and follow-up — ensuring operational efficiency and better decision-making.

At CarePro, we combine technology, data, and healthcare workflows to create measurable impact for hospitals and care teams.

The core platform is already live on MERN. Now we're upgrading the product with AI features users actually feel.

If you want to tinker with prompts forever, this isn't it.

If you want to ship, own outcomes, and build production-grade AI systems—let's talk.

You will

• Ship AI features for CarePro that solve real user pain (not toy demos)
• Own RAG end-to-end: ingestion → retrieval → outputs → evals → iteration
• Build with structured outputs + deterministic fallbacks (reliability > vibes)
• Monitor quality, latency, and cost—and improve it weekly
• Improve engineering velocity with AI workflows (Cursor/Claude/CodeRabbit style)

You should be

• A strong engineer first (APIs, debugging, testing, clean architecture)
• Comfortable with ambiguity + fast execution
• Opinionated about simplicity (agents only when they're worth it)
• Obsessed with measurable quality (eval harnesses + regression tests)

Must-haves

Shipped one AI/LLM feature (or built one to production-grade standards)

Solid with RAG + embeddings + retrieval strategies

• Strong in Node.js/TypeScript (MERN familiarity is ideal)

First 30 days = success

• Ship your first AI feature into CarePro production
• Put eval + regression in place so quality doesn't drift
• Leave behind a scalable pattern the team can reuse

Key Responsibilities

Build AI features into the platform

• Design and ship AI-driven capabilities (assistants, copilots, smart search, summarization, recommendations, extraction, classification, etc.)
• Implement LLM orchestration patterns (tool calling, structured outputs, multi-step workflows, agent loops when appropriate)
• Build and maintain RAG pipelines (chunking, embeddings, vector DB, retrieval strategies, reranking, citations)

Production-grade AI engineering

• Create evaluation frameworks (offline + online): quality scoring, regression tests, golden datasets, and A/B tests
• Implement guardrails: prompt hardening, safety rules, PII redaction, rate limiting, fallback logic
• Track performance with observability (latency, cost, accuracy, failure modes)

Improve engineering workflows with AI

• Enhance internal AI workflows for devs (PR review automation, test generation, refactoring assistants, doc generation, linting/quality gates)
• Standardize best practices for using Cursor/Claude/CodeRabbit-style tools in a consistent engineering system
• Build reusable templates and internal libraries for prompts, tools, and evals

Collaborate across teams

• Work with product and engineering to convert business problems into AI system designs
• Communicate tradeoffs clearly (accuracy vs cost vs latency, build vs buy, agent vs deterministic workflows)

Required Qualifications

• Strong software engineering fundamentals (clean architecture, API design, testing, performance)

• Experience shipping at least 1–2 AI/LLM features into production

• Solid knowledge of:

Prompting patterns + structured outputs (JSON schemas, function/tool calling)

RAG design and tuning (chunking, retrieval, reranking)

Evaluation methodologies (automated + human-in-the-loop)

• Hands-on development in Node.js / TypeScript and ability to work with React + MongoDB systems

• Familiarity with Git workflows, code review culture, and CI/CD

Preferred Qualifications (Nice to Have)

• Experience with vector databases (Pinecone, Weaviate, Milvus, FAISS) or Elastic hybrid search
• Experience with LangGraph / LangChain / LlamaIndex or similar orchestration frameworks (or equivalent custom systems)
• Background in ML fundamentals (embeddings, similarity search, classification, fine-tuning concepts)
• Experience with cloud deployment (AWS/GCP/Azure), Docker, Kubernetes (optional)

Experience setting up cost monitoring and optimization for LLM usage

Tech Stack

• Platform: MERN (MongoDB, Express, React, Node.js)
• AI stack (examples): OpenAI/Anthropic models, embeddings + vector DB, orchestration frameworks, evaluation harness
• Dev workflow tools: Cursor, Claude Code, CodeRabbit (and other AI-assisted coding/review tools)