Advancing Precision Medicine
through Open Source
An initiative dedicated to creating RWE and Precision Medicine solutions by integrating and developing open-source technologies across Clinical Informatics, Bioinformatics, Cloud Infrastructure, and MLOps.
Our core contributors bring academic and research backgrounds from:
Bridging healthcare innovation across California, Florida, and Spain.
Open Source Projects
ColabBio Base Platform
The foundational open-source framework powering the ecosystem. Designed for scalable, reproducible research with built-in Multi-Tenant isolation and Federated Learning capabilities.
Multi-model AI Pathology Viewer
An advanced visualization tool for digital pathology. Supports real-time multi-model AI inference for slide analysis, annotations, and biomarker detection.
Medical FUSE VFS
A unified virtual file system (VFS) based on FUSE, providing seamless native file access to diverse medical data stores including OMOP, FHIR, DICOM, and Genomics standards.
Enterprise MLOps & Inference
GPU-accelerated dynamic inference using NVIDIA Triton and MLflow. Features transparent NVIDIA NVFlare integration for federated learning across custom model dockers.
Clinical MCP Servers
Standardized medical Model Context Protocol microservices for OMOP CDM, DICOMweb, OMERO, and FHIR. Connects clinical data to AI agents safely over SSE.
Clinical Companion & A2UI
Agentic clinical copilot with 19 deterministic medical tools. Streams real-time Server-Driven UI (heatmaps, ROI focus, cards) directly into the viewer.
SlideLab Preprocessing
High-performance computational pathology pipeline for WSI preprocessing. Masks, normalizes, and tiles gigapixel images for downstream AI training.
Enterprise-Grade Architecture
ColabBio is built from the ground up for strict data governance, horizontal scalability, and privacy-preserving AI.
Federated Learning & Active Loop
RWE on Edge. Algorithms travel to the hospital data, ensuring 100% privacy. Continuous learning loops assist pathologists with real-time diagnostic overlays.
Learn more →Multi-Tenant Isolation
Strict namespace and database isolation. Host multiple departments on a shared cluster, or deploy physically isolated edge nodes in a zero-trust network.
Learn more →Flexible Deployment
3 Levels of Deployment: from local Docker Compose for rapid testing, to on-premise MicroK8s hospital clusters, up to scalable Cloud IaC (Terraform).
Learn more →Precision Medicine & Translational Pipeline
End-to-end support for digital health innovation, covering translation, RWE on edge, and clinical validation. We help Startups, Pharma companies, and Research Centers with their translational needs across Europe and LATAM.
Definition
Clinical guidelines, use cases, and technical requirements definition.
Piloting, Edge Training & Validation
Feasibility studies, federated learning, and algorithmic fine-tuning in controlled environments.
Certification
Regulatory compliance (FDA/CE), security audits, and interoperability certification.
Deployment
Full-scale commercial deployment, EMR integration, and continuous monitoring.
Open to Research Partnerships
We are actively seeking research centers, hospitals, and academic institutions to collaborate on public grants, consortiums (e.g., Horizon Europe), and open-source scientific pilots.
Collaborate with UsProfessional & Commercial Services
For professional deployments, proprietary EMR integration, and services outside the scope of open research, please visit ColabBio.net, an initiative driven by HealthCentrix.
Visit ColabBio.netSoonTenant Pilot: SlideLab/abMiL
A flagship computational pathology project demonstrating the power of ColabBio's open-source architecture.SlideLab/abMiL aims to develop robust AI models capable of identifying cancerous regions in Whole Slide Images (WSIs) without relying on expensive, pixel-level annotations.
By leveraging Attention-Based Multiple Instance Learning (ABMIL) and extracting features using foundation models, the model learns purely from weak slide-level labels.
Read Technical Docs