Celli AI — Intelligent Blood Smear Analysis Platform

    Celli AI is a research-driven and product-oriented project developed in collaboration with Shiraz University of Medical Sciences, bringing together a dedicated development team, AI specialists, and expert hematologists.

    Visit Project
    Technology Stack
    Django
    MySQL
    n8n
    OpenAI
    PostgreSQL
    Python

    The goal of this project was to build a Persian-language simulation and platform comparable to international Celli AI solutions, designed to make blood smear analysis in laboratories faster, more accurate, and more intelligent.

    Core Idea & Problem Statement

    In many laboratories, blood smear analysis is still performed manually and heavily dependent on human expertise — a time-consuming process that is highly sensitive to human error. Celli AI was created to address this challenge by:

    Leveraging artificial intelligence to analyze microscopic blood smear images and convert them into accurate, reliable, and actionable medical results.

    System Architecture & User Experience Celli AI was designed as a subscription-based platform tailored specifically for laboratories.

    Each laboratory could purchase a monthly or annual subscription and provide controlled access to its internal users.

    The system workflow was designed as follows:

    Each laboratory technician is provided with a dedicated user account. Users can log in to their personal dashboard and upload blood smear images.

    Images are automatically routed to a server running the AI analysis engine. The AI processes the images and generates a structured result file.

    The AI processes the images and generates a structured result file.

    The AI model developed for this project is capable of: Identifying and analyzing red and white blood cells. Evaluating cellular ratios and abnormal patterns. Detecting potential current medical conditions. Estimating future disease risks based on observed indicators. The generated results: Are first reviewed and validated by the laboratory technician, upon confirmation, are sent to the laboratory’s primary management system, used by administrative staff. This human-in-the-loop validation layer played a critical role in ensuring trust and adoption within clinical environments.

    Academic & Interdisciplinary Collaboration

    One of the defining strengths of Celli AI is its close interdisciplinary collaboration, involving: PhD-level artificial intelligence specialists, Medical doctors and hematology experts, Software engineers and system architects. This collaboration ensured that the AI models were not only technically sound but also clinically credible and medically validated.

    Technical Implementation

    Frontend

    Front-End: Built using React with TypeScript to deliver a secure, maintainable, and scalable user interface. Styled with Tailwind CSS to achieve a clean, professional, and enterprise-ready design.

    Backend

    Back-End: Powered by Django, responsible for core business logic, user management, subscriptions, and API services. Artificial Intelligence. AI models were fully developed in Python. Model design, training, and validation were carried out with the direct involvement of a PhD AI specialist from Shiraz University of Medical Sciences

    Final Outcome

    Celli AI is an intelligent medical platform that: Enhances the speed and accuracy of blood smear analysis. Reduces human error in laboratory diagnostics. Moves laboratory workflows toward AI-assisted medical automation

    This project stands as a strong example of how software engineering, artificial intelligence, and medical science can be combined to solve real-world, high-impact healthcare challenge.