Join Our Team at Compound

At Compound, we're leading the AI transformation revolution. We partner with mid-market and enterprise businesses to adopt AI in meaningful, impactful ways that drive real business value.

Position Overview

We are seeking a talented Applied AI Engineer to join our team for an exciting project: developing EMR Intelligence, an advanced AI-powered search and automation platform for large medical practice groups. This role involves building sophisticated natural language search capabilities across multiple healthcare data sources and EMR systems, plus developing intelligent agents that can execute complex tasks autonomously.

Project Summary

EMR Intelligence is a unified search and automation platform that enables healthcare professionals to query their entire data ecosystem using natural language. The system integrates with multiple EMRs, practice management systems, billing platforms, and other healthcare tools to provide instant, contextual answers. Beyond search, you'll build AI agents capable of executing multi-step tasks like scheduling, documentation, reporting, and workflow automation—transforming how medical practices operate.

Key Responsibilities

Search & Retrieval Systems

• Design and implement semantic search infrastructure across disparate healthcare data sources
• Build vector databases and embedding pipelines for efficient similarity search
• Develop hybrid search systems combining semantic and keyword-based approaches
• Create query understanding systems that interpret medical terminology and context
• Implement retrieval-augmented generation (RAG) pipelines for accurate, grounded responses
• Build ranking and re-ranking systems to surface most relevant results
• Optimize search performance for sub-second response times across millions of records

Natural Language Processing

• Develop NLP models fine-tuned for medical and healthcare contexts
• Build query parsing and intent classification systems
• Implement entity extraction for patients, providers, medications, procedures, and diagnoses
• Create disambiguation systems for medical terminology and abbreviations