BloomBridge

Empowering specially-abled children through love, care, and AI while ensuring every child’s safety with smart, compassionate technology.

Comment

Focus Area

-Healthcare & Well-being – AI for mental well-being, disease identification, and accessibility.

-Social Good & Humanitarian Impact – AI solutions that improve lives for underserved communities.

Inspiration

As a mother, when my son was diagnosed with mild ADHD in 2022, I was overwhelmed and unsure how to support him. Speech therapy was essential but incredibly expensive and not easily accessible. I searched for digital solutions, but most were either too complex or didn’t truly address our needs.

That’s when I realized in this era of AI, we can create something better.

BloomBridge is my answer a user-friendly, affordable platform built with love and purpose. It’s so intuitive that even women with limited digital literacy can use it confidently in their daily lives. It empowers caregivers, supports children emotionally and developmentally, and brings peace of mind through smart, compassionate technology.

As a developer, I believe we hold the power to create meaningful change and this is my way of giving back to society.

What it does

-BloomBridge is an AI-powered platform designed to support specially-abled and autistic children by providing emotional insights and personalized care.

-It leverages technologies like Amazon SageMaker for AI-driven emotional analysis, Amazon Polly for natural-sounding speech, and AWS Lambda with API Gateway for real-time child location tracking and safety alerts.

-BloomBridge aims to enhance communication, emotional understanding, and safety for these children, helping caregivers provide compassionate and effective support while ensuring children’s well-being and protection.

How we built it

Frontend Development
We used HTML, CSS, JavaScript, and Leaflet.js for interactive maps. The interface is simple and accessible with calming blue tones for a sensory-friendly experience. The app is deployed on GitHub Pages for fast, free, and accessible delivery.

Core Functionalities & AWS Integration
-The Speech Therapy Module is built using Amazon Polly. It converts text into natural-sounding speech to guide children through words or routines and helps simulate occupational and speech therapy sessions.

-Child Location Tracking and Panic Alert use the browser’s geolocation API to fetch the child’s real-time location. When the Panic Button is clicked, coordinates are sent to Amazon API Gateway, processed by an AWS Lambda (Node.js) function, which calls LocationIQ API to get the address and sends a notification via Amazon SNS to the caregiver’s phone or email.

-Daily Routine and Picture-to-Word features use interactive visuals with click-to-play functions. Text and image pairing helps boost comprehension and memory.

-Emotion Logging allows caregivers to enter text describing the child’s behavior. In future, this data will be processed using Amazon SageMaker JumpStart to classify emotions like anger, joy, or frustration.

Project Walkthrough

Refer this PDF

Refer this realtime demo

Challenges we ran into

-We faced CORS issues on API Gateway, which we fixed with proper OPTIONS methods and headers. We gained an understanding of how SNS topics and subscriptions work. We improved our understanding of AI, and learned to use LeafletJS and geolocation together. Overall, we gained insights into serverless application design.

-Autistic children may express themselves differently, and caregivers' descriptions can vary widely. The generic emotion classification models occasionally misinterpreted context labeling overwhelmed behavior as "anger" or missing signs of sensory overload.

-Deploying real-time generative AI models, even using SageMaker JumpStart, incurs costs. Since BloomBridge is designed to be free and accessible, balancing model performance and affordability was a significant challenge.

Limitations

  • Amazon SageMaker Cost Constraint:
    While the application was initially designed to use SageMaker for real-time emotion detection, cost limitations during development led to a shift toward collecting caregiver input and providing emotional support responses based on that input. This approach still supports the core goal of helping children but in a more cost-effective manner.

  • Live Deployment Constraints:
    Some features (e.g., emotion classification via SageMaker) may be limited or simulated due to free-tier restrictions.

Accomplishments that we're proud of

-One of the key accomplishments of BloomBridge is its grounding in real-world insights. We conducted a case study involving 10–15 caregivers of specially-abled and autistic children to understand their daily struggles and unmet needs. Their feedback directly shaped the core features of the app such as speech therapy powered by Amazon Polly, daily routine management, and live location tracking.

-The app was tested in real caregiving settings, showing promising results with improved child engagement and caregiver satisfaction. BloomBridge stands out for being affordable, user-friendly, and designed for accessibility, especially for parents who may not be tech-savvy.

-The result is a solution that makes a real difference happier children and empowered caregivers.

Refer this Case study form

What we learned

-Through building BloomBridge, we learned the power of combining empathy with technology. Caregivers want simple, accessible tools to help their children communicate, feel safe, and grow.

-We discovered that even lightweight AI tools like Amazon Polly can make a big impact when applied with real-world feedback. Our survey of 10–15 mothers helped us prioritize features like speech therapy, daily routines, and child tracking.

-We also realized the importance of intuitive design, especially for parents with limited tech exposure. Most importantly, we learned that thoughtful tech can bring comfort, independence, and joy to specially-abled children and their families.

What's next for BloomBridge

BloomBridge will use advanced AWS services like Amazon SageMaker JumpStart for emotion detection, Amazon Polly for speech synthesis, and Amazon Rekognition for visual cues.

-We’ll also add interactive features using JavaScript and modern web frameworks, ensuring privacy and accessibility for all users.

Built With

Share this project:

Updates