Reasoning AI + Personalized Clinical Conversations

RedcellAI analyzes CBC/CMA to detect hidden anomalies (FN=0 target), explains why with SHAP, and opens a bilingual, role-aware conversation for doctors and patients — from insights to action.

🚀 Our Vision

RedcellAI is building the world’s first Reasoning AI for Healthcare with a personalized conversational mode for both doctors and patients. We don’t stop at anomaly detection — we generate role-aware, explainable dialogues that help clinicians decide faster and patients act earlier.

🎥 Watch Our Demo

See how RedcellAI turns routine labs into explainable, conversational insights.

💬 Personalized Interaction Mode

Doctor Mode: SHAP-linked differentials, red-flag cues (FN=0 target), guideline snippets, PDF export.
Patient Mode: plain-language summaries, next-step guidance, lifestyle tips, bilingual (KO/EN).
Safety: guardrails, uncertainty cues, feedback logging.

Role-aware prompts (Clinician ↔ Patient)
Question generation tied to SHAP features
Multilingual output (KO/EN) & reading-level control
Feedback → continuous learning → better conversations

🧠 Why We’re Different

Reasoning (not just scores) + FN=0 anomaly engine + Gemini-powered dialogue = measurable clinical impact.

Ensemble outlier detection (IF/AE/LOF)
SHAP explanations → questions → actions
Doctor-grade outputs (differentials, red flags)
EMR/PDF/API integration

News & Updates

May 2025: 🎉 RedcellAI approved for the NVIDIA Inception Program. 📄 View Approval
May 2025: 🎉 RedcellAI selected for the Google for Startups Cloud Program, receiving $25,000 in credits. 📄 View Approval

RedcellAI Blog

🧠 Pilot 4.0: Ensemble AI for Blood Test Interpretation

We’ve launched Pilot 4.0 using Isolation Forest, XGBoost, AutoEncoder, SHAP, and Gemini Pro for explainable diagnostics.

🧠 We Built an AI That Reads Blood Tests Like a Hematologist

Discover how we built an AI assistant that not only detects hematologic abnormalities, but also reasons and explains like a hematologist — using SHAP, Gemini, and zero-tolerance outlier detection.

👉 Read full article →

Product Introduction

RedcellAI combines unsupervised learning (IF, LOF, AE) with generative AI (Gemini) to analyze blood tests, detect anomalies, generate reports, and enable interactive Q&A.

Request a Demo

We’re now accepting beta pilot applications from clinics and institutions. Apply here.

Pricing

Plan	Description	Price
Starter Clinic	Up to 300 tests/month, auto reports	$99/month
Pro Clinic	Up to 1,000 tests/month, SHAP explainability	$249/month
API Plan	JSON API for EMRs/Insurers	$0.20/test

* 3-month free pilot available for Korean clinics during beta.

Our Team

Jubeam Lee (CEO) – AI product & startup veteran.
Hyun-Ju Seo (CTO) – SaaS & ML operations expert.
Dr. Jaewoo Song (Clinical Advisor) – Professor, Yonsei University Hospital.

Powered by RedcellAI

🔬 91.2% accuracy in CBC-based disease detection
📊 SHAP & Gemini for model interpretability
🤝 Collaboration with Yonsei University Severance Hospital

FAQ

What data does RedcellAI use?
CBC, CMA, and clinical chemistry panels.

Can I try it with real data?
Yes — upload sample CBC results in our demo.

Is the platform secure?
Data encrypted in transit & at rest; HIPAA-ready.

Contact Us

Email: [email protected]