Turning Complex Medical Journeys into Clear Insights.
A patient-centric health intelligence platform that transforms your medical records into personalized, literacy-adapted summaries — so you can understand, prepare, and advocate for your care.
What We Do & Why It Matters
Health literacy is a silent crisis — and we're built to close the gap.
The Problem
Patients leave appointments confused — and that confusion is dangerous.
An estimated 36% of U.S. adults have limited health literacy. [NCES NAAL ↗] Yet patients are routinely handed dense discharge summaries and medication lists filled with clinical jargon they are never equipped to interpret.
The result: patients miss critical warning signs, fail to adhere to treatment plans, and are ill-prepared to advocate for themselves. For those managing multiple chronic conditions simultaneously, the information gap compounds with every appointment.
Meanwhile, EMRs contain rich, actionable data about a patient's health — but that data is locked inside provider systems, never synthesized for the person it concerns most.
West Virginia ranked 47th out of 50 states in health system performance
Commonwealth Fund Scorecard, 2025 ↗of WV adults have diabetes — the highest rate in the nation
America's Health Rankings — WV Diabetes ↗of WV adults report fair or poor health — ranked 2nd highest in the nation
CDC Behavioral Risk Factor Surveillance System ↗Our Solution
Personalized health intelligence, in plain language.
Access Insight connects to a patient's medical profile and runs it through a proprietary multi-specialist AI architecture — 16 domain-specific AI agents (cardiologist, endocrinologist, neurologist, pharmacist, and more) consult in parallel, exactly as a multidisciplinary care team would.
The results are synthesized into a personalized, reading-level-adapted health dashboard grounded in validated health literacy scoring (NVS + BHLS). Every summary is calibrated to the individual — not a generic patient population.
Patients can then chat with an AI companion that already knows their full medical context, deep-dive into any term for instant plain-language explanations, and walk into every appointment prepared with the right questions.
How Access Insight is Different
-
Literacy-adaptive output — summaries are rewritten based on a validated literacy score, not a one-size average.
-
Multi-specialist reasoning — 16 AI specialists surface cross-condition interactions a generalist model would miss.
-
MedlinePlus integration — every condition and medication links to authoritative NIH-curated resources.
-
Context-aware chat — the AI companion already has your full health profile, so every answer is relevant to you.
-
Patient-owned — unlike EMR portals, insights are delivered directly to the patient, in their language.
Built in Appalachia — For Every Underserved Patient
Access Insight was created at Marshall University in West Virginia — a state that ranks among the most medically underserved in the nation. West Virginia has the highest diabetes rate in the U.S. (18.4%), [AHR ↗] the 2nd highest cardiovascular disease prevalence in the nation (14.2%), [AHR ↗] and the highest COPD rate in the nation (13.6%) [CDC ↗] — all conditions generating complex, difficult-to-understand medical documentation.
The Commonwealth Fund's 2025 State Health System Performance Scorecard ranks West Virginia 47th out of 50 states nationally. [Commonwealth Fund, 2025 ↗] Over 26% of WV adults report fair or poor health — ranked 2nd highest in the nation. [CDC BRFSS ↗] This is the patient population we know firsthand, and whose needs shaped every design decision in Access Insight.
Target Audience
Primary: Adults managing one or more chronic conditions — including diabetes, cardiovascular disease, autoimmune disorders, and complex medication regimens — who struggle to interpret their medical information.
Secondary: Patients with limited formal education or English as a second language, who face amplified barriers to health understanding.
Institutional: Health systems, federally qualified health centers (FQHCs), and health equity initiatives seeking scalable patient education tools.
Industry Context
Access Insight sits at the intersection of digital health, AI-assisted clinical decision support, and health equity. The global digital health market is projected to exceed $800B by 2030, driven by AI adoption and a shift toward patient-centered care models.
Recent federal mandates (21st Century Cures Act) require providers to give patients immediate access to their health data. Access Insight makes that data understandable — the critical missing layer.
Our technology leverages Google Gemini 2.5 Flash on Google Cloud Run, with a proprietary multi-agent architecture that differentiates us from generic consumer health AI chatbots.
Your Path to Clarity
Three steps from medical complexity to confident understanding.
Share Your Profile
Securely enter your medical history, conditions, and prescriptions. No raw EMR files required — just the information you already know about your health.
AI Specialist Consult
16 domain-specific AI agents — cardiologist, endocrinologist, pharmacist, and more — review your profile in parallel, simulating a multidisciplinary care team consultation.
Gain Clarity
Receive a personalized, reading-level-adapted dashboard. Chat with your health companion, click any term for a Deep Dive explanation, and walk into every appointment prepared.
The People Behind Access Insight
Founded and built by medical students who experience the patient communication gap firsthand — every day on the wards.
Why We Built This
During clinical rotations at Marshall University Joan C. Edwards School of Medicine, we watched patients leave appointments holding discharge papers they couldn't read, unable to explain their own diagnoses. We knew AI could close this gap — but only if it was built with the patient, not the provider, as the primary user. Access Insight is the tool we wished existed for every patient we saw.
Bukola Adeshina
Co-Founder & CEO
MPH, Johns Hopkins University | MD Student, Marshall University JCESOM
- Master of Public Health — Johns Hopkins Bloomberg School of Public Health
- MD Candidate — Marshall University Joan C. Edwards School of Medicine
- Public health background informs the platform's health equity focus and literacy-adaptive design
- Leads clinical validity, specialist agent accuracy, and patient experience
Joshua Eastman
Co-Founder & CTO
BSE, Princeton University | MD Student, Marshall University JCESOM
- Bachelor of Science in Engineering — Princeton University
- MD Candidate — Marshall University Joan C. Edwards School of Medicine
- Designed and built the full-stack platform: FastAPI backend, multi-agent AI architecture, Flutter mobile app, and GCP cloud infrastructure
- Engineering background enabled end-to-end construction of the platform alongside medical training — no external engineering team
Clinical Founders
Built by medical students with direct patient contact — not just technologists
Google Cloud
Platform runs on GCP Cloud Run with Gemini 2.5 Flash at its core
Evidence-Grounded
Literacy scoring uses validated NVS + BHLS clinical instruments
What We're Building
Currently in Early Access Beta — live on Google Cloud Run, actively used by beta participants.
MVP
Core dashboard, multi-specialist AI, literacy assessment, and chat — complete.
Early Access Beta
Live platform with real users. Collecting feedback, refining UX, expanding features.
General Launch
Expanded access, institutional partnerships, and mobile app public release.
Core Feature Set
Personalized Health Dashboard
AI-generated summary of conditions, medications, interactions, lifestyle guidance, and warning signs — adapted to the patient's reading level. Structured into tabbed sections: My Health, Learn, Prepare, Ask.
Multi-Specialist AI Architecture
16 domain-specific AI agents — Cardiologist, Endocrinologist, Neurologist, Pharmacist, Rheumatologist, OB/GYN, and more — run in parallel using Google Gemini 2.5 Flash with MedlinePlus retrieval.
Context-Aware AI Chat
A conversational health companion that already knows the patient's full medical profile and dashboard context. Supports follow-up questions, history saving, pinning, and folder organization.
Deep Dive Explanations
Click any medical term to instantly receive a plain-language explanation calibrated to the patient's literacy level, with MedlinePlus citations and built-in follow-up exploration.
Health Literacy Assessment
Integrates the Newest Vital Sign (NVS) and Brief Health Literacy Screen (BHLS) to score literacy as Limited, Marginal, or Adequate — dynamically adjusting all AI outputs accordingly.
Health Journal
A personal health log for recording symptoms, questions, and observations between appointments — creating a continuous record that complements the AI dashboard.
Cross-Platform Mobile App
Flutter-built iOS and Android app with biometric authentication, offline caching, and the full dashboard experience optimized for mobile. Available to beta users.
Privacy & Security
Firebase Authentication, HTTPOnly session cookies, PostgreSQL with pgvector on GCP, and CI/CD via Cloud Build. Patient data is encrypted at rest and in transit.
MedlinePlus Integration
Every condition and medication card is enriched with NIH MedlinePlus data — authoritative plain-language summaries and links — giving patients a trusted research path beyond AI-generated content.
Technology Stack
AI / ML
- Google Gemini 2.5 Flash
- Multi-agent function calling
- MedlinePlus SQLite RAG
- pgvector (PostgreSQL)
Backend
- Python / FastAPI
- SQLAlchemy + Alembic
- PostgreSQL (Cloud SQL)
- Firebase Auth
Frontend & Mobile
- Jinja2 + Bootstrap 5
- Vanilla JS (IIFE modules)
- Flutter (iOS + Android)
- Riverpod state management
Infrastructure
- Google Cloud Run
- Cloud Build CI/CD
- Docker
- Google Cloud SQL
Near-Term Roadmap
Dashboard UX Overhaul
Tab-based navigation, enhanced specialist cards, and richer MedlinePlus surfacing.
Prepare for Visit
AI-generated pre-visit prep, appointment logging, and question prioritization.
USPSTF Screening Alerts
Proactive preventive care reminders based on patient age, sex, and conditions.
Institutional Pilot Partnerships
Deploying with FQHCs and academic medical centers as a patient education layer.
Get In Touch
We are actively seeking pilot partners, feedback from clinicians, and conversations with organizations aligned with health equity and patient empowerment.