AlphaGrid AI - Oncology Intelligence Platform Development

An oncologist preparing for a patient consultation faces a nightmare: imaging reports in the PACS system, genomic data in the LIMS platform, treatment history in the EMR – all separate systems requiring multiple logins. Piecing together a complete patient picture takes 45+ minutes per case.

Then AI tools promised to help, but doctors rejected them. Why? Without source transparency, AI in cancer care was useless

Case prep time: 45+ minutes → under 10 minutes

AI adoption increased through transparent, traceable insights

Complete patient context eliminates missed critical information

Legacy systems integrated without infrastructure replacement

Faster diagnosis and treatment planning (days saved per case)

No more repeating medical history at every appointment

Reduced medical errors from incomplete data

Oncology intelligence platform providing traceable AI insights for cancer care, centralising patient data from PACS imaging, LIMS genomics, and EMR systems.

My Role: Frontend developer and AI traceability architect

Tools Used: Cursor, Figma, React DevTools

Tech Stack: React.js, Node.js, PostgreSQL, AWS S3, DICOM viewers, OpenAI API

Unified Patient Dashboard

AI Summarization Engine

Real-Time Granular Traceability System

Unified Reporting System (URS) Generator

DICOM Medical Image Viewer

Role-Based Access Control (Radiologists, Doctors, Admins)

Historical Data Aggregator

Citation & Source Linking

PACS systems for imaging data

LIMS for genomic data

EMR systems for patient records

DICOM image processing

OpenAI for AI summarization

Doctors don’t trust AI without source transparency (built hover-to-cite traceability)

Oncology data is siloed across systems (unified longitudinal view)

DICOM images are complex to render (custom viewer implementation)

Clinicians need fast insights without losing detail (AI summarization with deep-dive capability)

Legacy systems don’t integrate easily (built custom connectors)