top of page

Project ID:

ITCS04313

Caring

Project Title:

Category:

Information Technology/ Computer Science/ AI

Inventors:

Louis Tey Chi Feng, Eric Yeong Hong Wei, Lee Kai Siang, Ts. Shaffika

Institution/Company:

Southern University College

Invention Description/ Abstract:

Modern dual-income families face substantial challenges in securing reliable and flexible childcare due to rigid daycare operating hours, unforeseen overtime, and trust deficits inherent in conventional domestic agencies. At the same time, independent caregivers often experience fragmented compensation, unclear expectations, and a lack of formalized labor protections. Caring is a mobile cloud platform designed to address these challenges by establishing a trusted, transparent, and intelligent caregiving ecosystem.

Developed using Flutter and backed by Firebase cloud infrastructure, Caring integrates several key technical and operational innovations. The platform incorporates an AI-powered semantic matchmaking engine using large language models through the Groq LLaMA-3 API. It transforms open-ended parental requests, such as newborn care, meal preparation, or specialized tutoring, into multidimensional parameters that consider caregiver skills, hourly budgets, geographical location, and historical ratings. Based on these factors, the system provides suitable caregiver matches together with natural-language explanations.

To enhance household safety, Caring also implements a biometric electronic Know-Your-Customer (KYC) verification system. Caregivers are required to complete identity verification and live facial detection before they can activate their "Online" status and receive booking requests. The platform further improves service transparency through a real-time care milestone timeline. During an assignment, caregivers can record structured updates such as feeding, nap times, and interactive activities, accompanied by photographic timestamps. At least one validated care update is required before a booking can be completed and payment can be released.

In addition, Caring features an intelligent slot-based scheduling system that automatically detects and prevents overlapping bookings. A mandatory one-hour rest buffer is enforced after each confirmed assignment to reduce caregiver fatigue and support consistent service quality. By directly aligning with United Nations Sustainable Development Goals 3 (Good Health and Well-being) and 8 (Decent Work and Economic Growth), Carin

Invention Technical Description

Caring is a cross-platform mobile application designed to address security concerns, matchmaking inefficiencies, and welfare issues in on-demand childcare. The application is developed using the Flutter SDK, enabling cross-platform deployment and consistent user interface rendering. Google Cloud Firebase is used as the backend infrastructure to provide cloud database storage, real-time data synchronization, and user authentication.

The system consists of four main technical components. First, Caring uses an LLM-based semantic matchmaking engine integrated with the Groq Cloud API and powered by LLaMA-3. When parents enter their childcare requirements in natural language, the system processes the information and converts it into structured parameters, including required skills, hourly budget, and geographical preferences. These parameters are then compared with caregiver profiles stored in the cloud database using a multi-criteria matching algorithm. Suitable caregivers are ranked according to the matching results, while the system also provides a simple explanation of why each caregiver is recommended.

Second, Caring implements an electronic Know-Your-Customer (KYC) verification process to improve platform security. Caregivers are required to complete identity verification and a front-camera facial detection procedure before they can provide services through the platform. The verification status is stored in the Firestore database, and security rules prevent unverified caregivers from changing their status to "Online" or receiving booking requests until the required verification has been completed.

Third, the platform provides a real-time care milestone timeline to improve transparency between caregivers and parents. During an active booking, caregivers can record important activities such as meals, nap times, and safety checks, together with timestamps and photographic evidence. The system requires at least one validated care milestone before a booking can be marked as completed and the payment can be released. This provides parents with an activity record that allows them to monitor the care provided during the booking.

Finally, Caring incorporates a slot-based scheduling system to prevent booking conflicts and reduce caregiver fatigue. Once a booking is confirmed, the system automatically adds a mandatory one-hour rest period after the scheduled care session. Any new booking that overlaps with either the active care period or the required rest period is automatically identified and rejected, while the affected parent receives a notification. This scheduling mechanism helps maintain a more organized booking process while providing caregivers with sufficient recovery time between assignments.

Demostration/ Presentation Video

Poster/ Broucher/ Invention Photo

Additional Documents

> Additional Technical Document
> Additional Photo of the Invention

Copyright ©2026 by I-RICE 2026. All rights reserved

bottom of page