Report Samjho

Landing Page
Features and How it works
Pricing
Authentication
Report Upload
RAG Health Consultation Chat
Health Risk Timeline and Treads
Multi-Language and Voice Readout
AI Report Analysis

About the Project

The Problem That Inspired Us

Medical reports are written for healthcare professionals—not ordinary patients.

Terms such as hemoglobin, bilirubin, creatinine, reference range, and differential count can be difficult for a patient to understand. Even when a report marks a value as high or low, patients often do not know what it means, how important it may be, or what questions they should ask their doctor.

The problem becomes even bigger when patients are more comfortable in a language other than English.

We wanted to bridge this communication gap without replacing doctors or providing unsafe diagnoses.

That led us to ReportSamjho AI — an AI-powered system designed to transform complex medical reports into simple, understandable, multilingual explanations while preserving the original medical context.

What We Built

ReportSamjho AI takes a patient's medical report and converts its complex information into an easy-to-understand explanation.

The system is designed to:

Extract important information from medical reports.
Identify relevant biomarkers and their reported values.
Compare values with the reference ranges provided in the report.
Explain medical terminology in simpler language.
Highlight potentially concerning values for patient awareness.
Provide contextual explanations rather than isolated definitions.
Support multilingual understanding.
Ground explanations in trusted medical information using a retrieval-based approach.
Clearly distinguish health information and interpretation from medical diagnosis.

Instead of simply saying:

“Hemoglobin: 10.2 g/dL — Low”

the goal is to provide an explanation that a non-medical user can actually understand, such as what hemoglobin represents, why the reported value is outside the provided reference range, and what questions they may discuss with a qualified healthcare professional.

How We Built It

Our approach combines document processing, natural language processing, retrieval-augmented generation (RAG), and multilingual AI.

A simplified pipeline is:

Medical Report → Information Extraction → Medical Context Retrieval → AI Reasoning → Simplification → Multilingual Explanation

The system is designed around a grounding principle: explanations should be generated using the information present in the patient's report and relevant trusted medical knowledge rather than relying solely on an unconstrained language model.

This helps reduce hallucination and keeps the generated explanation connected to the actual report.

What We Learned

Building ReportSamjho AI taught us that healthcare AI is not simply about choosing the most powerful model.

The harder problem is creating a system that is:

understandable to non-technical users,
grounded in reliable information,
multilingual,
transparent about uncertainty,
and careful not to cross the boundary between explanation and diagnosis.

We learned to think about AI as a communication and accessibility layer rather than simply a prediction engine.

Challenges We Faced

One of our biggest challenges was balancing simplicity with medical accuracy.

Oversimplifying a medical term can remove important context, while using medical terminology makes the explanation difficult for patients. We therefore focused on contextual explanations rather than direct word-to-word translations.

Another challenge was multilingual communication. Translating medical concepts is not enough; the explanation must remain understandable and preserve the intended medical meaning.

We also had to consider the risks of generative AI in healthcare. A system that confidently invents information can be harmful. This is why grounding, clear uncertainty handling, and responsible AI principles are central to our approach.

Why It Matters

A medical report should not become a wall of unfamiliar terminology between a patient and their own health information.

ReportSamjho AI aims to make medical information more accessible, understandable, and language-inclusive, helping patients better understand what their reports contain and communicate more effectively with healthcare professionals.

Our vision is not to replace doctors.

Our vision is to make sure patients understand what their doctors are talking about.

Future Scope

We envision expanding ReportSamjho AI beyond individual laboratory reports toward a broader patient health-information understanding platform, capable of working with different types of medical documents and longitudinal health records.

Future development could include stronger clinical validation, additional Indian and global languages, multimodal medical-document understanding, personalized longitudinal explanations, and integration with healthcare workflows under appropriate clinical and regulatory safeguards.