[Hiring] Full Stack AI Engineer @Caretria

Role Description

We’re looking for a Fullstack AI Engineer to build AI applications for healthcare workflows. You’ll develop retrieval-augmented generation (RAG) systems, agentic workflows, and real-time voice agents, taking features from prototype through production.

This role combines hands-on Python and JavaScript development with a strong understanding of LLM integrations, conversation state, and reliable application architecture. You’ll work closely with product and engineering teammates to deliver systems that are useful, measurable, and designed with sensitive healthcare data in mind.

What you’ll do

• Design and build RAG pipelines, including document ingestion, chunking, embeddings, vector storage, and retrieval quality testing.
• Integrate LLM provider APIs, including Azure AI Foundry, Amazon Bedrock, and OpenAI, balancing quality, cost, and latency.
• Develop agentic workflows using LangChain and LangGraph, including tool calling, multi-step orchestration, and persistent conversation state.
• Build real-time voice agents using WebSockets, streaming speech-to-text and text-to-speech pipelines, and platforms such as Retell.
• Develop application features and backend services using Python 3.12+ and JavaScript/Node.js.
• Design REST APIs and asynchronous processing workflows using message queues such as BullMQ, RabbitMQ, or Amazon SQS.
• Implement structured outputs, validation, retries, and fallback behavior to make AI features dependable.
• Establish tracing, evaluation, and monitoring to assess model behavior, diagnose failures, and improve performance.
• Containerize services with Docker and Docker Compose, and contribute to AWS deployments and CI/CD pipelines.
• Collaborate on sprint planning, code reviews, and technical decisions while applying privacy and security practices appropriate for healthcare applications.

Qualifications

• Strong practical experience building RAG systems, with a clear understanding of vector search, embeddings, chunking strategies, and retrieval tradeoffs.
• Hands-on experience integrating LLM provider APIs and shipping AI features into production.
• Proficiency in Python and JavaScript/Node.js, including API development and service integrations.
• Experience with LangChain and LangGraph for orchestrating stateful LLM workflows.
• Experience building real-time voice applications, including streaming, turn-taking, interruption handling, and latency optimization.
• Familiarity with message queues and reliable background processing.

Requirements

• Working knowledge of Docker, Docker Compose, Git, and GitHub.
• Familiarity with CI/CD tools such as AWS CodePipeline or equivalent.
• Experience working in an iterative development environment using Jira and sprint planning.
• Awareness of HIPAA and the responsibilities involved in handling protected health information (PHI).
• Clear communication, sound technical judgment, and the ability to turn loosely defined requirements into maintainable software.

Additional experience we value

• Parse and MongoDB.
• AWS services such as ECS, Lambda, and Application Load Balancer.
• Healthcare software or other applications involving sensitive data.
• Multi-agent coordination and long-running workflows with checkpointing, memory, and persistence.
• Prompt engineering, schema-constrained generation, and function/tool definitions.
• LLM evaluation, observability, and testing of non-deterministic systems.
• Mitigating hallucinations, prompt injection, and unsafe tool use through validation and guardrail design.

What success looks like

You’ll deliver AI features that perform reliably in real workflows, with measurable retrieval quality, responsive voice interactions, and clear visibility into system behavior. You’ll make thoughtful tradeoffs across model quality, cost, latency, and context limits while helping the team maintain secure, understandable, and production-ready software.