ThalassaPredict AI is an innovative, anonymous web-based tool that provides predictive probabilities of blood availability at various blood centers. It transforms reactive searching into proactive planning.
• Fact-Based Statistical Prediction: Our AI/ML models are trained exclusively on reliable, aggregated statistical data (numbers and facts) from official sources like blood banks, government reports, and e-RaktKosh. This ensures predictions are grounded in verifiable data, not personal information or speculative sources, fostering unparalleled trust.
• Absolute Anonymity: A core principle: no user login, no personal profiles, and no collection of any personal user data. This eliminates privacy concerns (e.g., DPDP Act 2023) and complex security burdens, making it a truly accessible public utility.
• Intelligent, Proactive, Location-Aware Recommendations: The system provides dynamic, actionable recommendations: ◦ Users first select if they need blood for their current location (radius defaults to town/city size) or another place (user inputs location and custom radius). ◦ The AI calculates the probability of the requested blood type being available in the selected area based on historical statistical trends. ◦ Conditional Recommendations: If the local probability is below 15%, OR if a blood bank within a 50km radius shows a probability exceeding the local probability by more than 40%, the system recommends exploring alternatives. ◦ If no blood bank in the initial 50km radius meets the criteria, it identifies the nearest possible blood bank (potentially beyond 50km) that does exceed the local probability by at least 40%. It then lists all other high-probability blood banks within that extended radius, sorted by probability.
• Hybrid Live & Predictive Data Display: For searches on the current date, AI predictions are seamlessly integrated with live, real-time blood availability data from e-RaktKosh APIs, offering the most comprehensive and up-to-the-minute information. For future dates, only the predictive probability is shown.
• Anonymous Feedback Loop for Data Reliability: Users can provide simple, anonymous feedback on the reliability of the displayed data (both predicted and real-time), which is used to continuously refine the underlying statistical models and validate real-time data sources.
