H
HealthLeap AI Tech Stack
AI screening platform for hospital clinicians to identify missed diagnoses
HealthLeap builds an AI screening system deployed directly into hospital workflows (Epic, Cerner) to surface undiagnosed conditions clinicians miss under workload. The stack reveals a data-heavy, ML-first engineering approach: Spark, Dagster, Airflow, scikit-learn, XGBoost, and LightGBM for batch and streaming pipelines processing millions of patient records daily, paired with LLM-based clinical-note parsing to extract signals at scale. Leadership hiring (8 senior+ roles) and a 4-person engineering team suggest early-stage infrastructure investment to solve real-time risk ranking and EHR integration — two pain points blocking both clinical and financial impact proof.
Tech Stack 25 technologies
Core StackReact TypeScript AWS Kubernetes Epic Systems Terraform Python pandas scikit-learn PostgreSQL Iceberg Apache Spark Dagster Apache Airflow Cerner SQL Polars XGBoost LightGBM HTML/CSS AWS ECS AWS EKS Jupyter
What HealthLeap AI Is Building
◆Challenges
- Lack of design team
- Conditions missed by overworked clinicians
- Low diagnosis rates for malnutrition
- Lack of real-time risk ranking
- Integrating poorly documented ehr systems
- Processing millions of patient records daily
- High concurrency systems
- New condition expansion
- Clinical impact proof
- Financial impact proof
▲Active Projects
- Data pipelines for millions of patient records
- Product for malnutrition
- Parse 50,000+ clinical notes per patient using llms
- Product for copd
- Screening system for hospital patients
- Clinical notes parsing with llms
- Launch new conditions
- Product for chf
- Deploy live models
- Malnutrition model improvement
Hiring Activity
Steady9 roles · 3 in 30d
Department
Seniority
Company intelligence
