Delirium Prevention AI App for In Hospital use
Budget: $1,500 – $3,000 CAD
Project Overview
We are working on a compassionate tech project aimed at supporting patients with
dementia. The system enables patients to speak to an iPad, where their speech is
transcribed and matched to pre-recorded video messages from loved ones. This project aims
to provide support for dementia patients by offering a comforting and responsive
application when loved ones aren’t present. The application is intended to allow patients to
speak directly to an iPad application, which transcribes their speech into text. This text is
then matched with the most appropriate response from a database, triggering a
pre-recorded video message from a loved one. This personalized interaction helps patients
feel connected and reassured, even when their loved ones aren't physically available.
Technical Details:
The project consists of a front-end and back-end architecture. The front-end is built using
Flutter and is intended to run on iOS devices, particularly iPads. The back-end leverages
Python and key libraries such as NLTK for natural language processing and Flask for
creating a lightweight web framework. It also includes a shallow neural network
implemented with Keras for text matching and response selection. The application
processes the transcribed text, matches it against a database of possible responses, and
plays the corresponding pre-recorded video message.
The codebase already exists and will be provided upon hiring.
We are seeking a freelancer to implement key enhancements and prepare the app for
real-world iPad use.
The project is time-sensitive, so bids with a shorter timeline will be prioritized.
Key Tasks
1. Automatic End-of-Speech Detection:
- Remove the need for the user to press a button to end speech input.
- Implement an automatic end-of-speech detection mechanism.
- Ensure smooth integration with the existing backend flow that triggers transcription
and video response.
2. App Packaging for iPad Use:
- Build and package the Flutter app into an installable iOS format (IPA or
TestFlight-ready).
- Ensure compatibility and ease-of-use for caregivers and patients with no technical
expertise.
- Handle any permissions or audio-related configurations needed for deployment on
iPads.
- Provide simple installation instructions or an onboarding guide for caregivers.
3. Optional Improvements (if time and skills allow):
- Refactor the code and add extensive documentation for future developers expected to
continue working on this project.
- Provide packaging instructions and document any problems with their solutions
Required Skills
- Python (Back End):
• NLP experience, especially with NLTK.
• Familiarity with neural networks and DL tools (Keras).
• Experience with Flask..
- Flutter (Front End):
• Strong experience in building cross-platform apps.
• Integration of speech/audio functionality in Flutter.
• State management and asynchronous operations.
- Speech & Audio Processing:
• Understanding of end-of-speech detection or VAD (voice activity detection) techniques.
• Experience with audio stream handling in mobile environments.
- iOS Deployment:
• Experience building and signing Flutter apps for iOS.
• Familiarity with Apple provisioning profiles, certificates, TestFlight, or direct IPA creation.
Deliverables
- Updated well-documented codebase.
- Working app with automatic speech end detection.
- iOS installable package ready for testing and use on iPads.
- README and short installation guide.
Timeline (Estimated Hours)
• Automatic End-of-Speech Detection: 20–30 hours
• Packaging and Testing for iPad: 15–25 hours
• Documentation and guide preparation 3-5 hours
Total Core Task Estimate: 38–60 hours
How to Apply
Please share your relevant experience, particularly with: Flutter + iOS packaging,
Speech/audio processing, and Apps designed for health or accessibility.
We’re excited to work with someone who shares our passion for tech with purpose. We look
forward to receiving your proposal with the estimated timeline and expected budget. We
will schedule meetings with prospective developers to finalize the details.
We are working on a compassionate tech project aimed at supporting patients with
dementia. The system enables patients to speak to an iPad, where their speech is
transcribed and matched to pre-recorded video messages from loved ones. This project aims
to provide support for dementia patients by offering a comforting and responsive
application when loved ones aren’t present. The application is intended to allow patients to
speak directly to an iPad application, which transcribes their speech into text. This text is
then matched with the most appropriate response from a database, triggering a
pre-recorded video message from a loved one. This personalized interaction helps patients
feel connected and reassured, even when their loved ones aren't physically available.
Technical Details:
The project consists of a front-end and back-end architecture. The front-end is built using
Flutter and is intended to run on iOS devices, particularly iPads. The back-end leverages
Python and key libraries such as NLTK for natural language processing and Flask for
creating a lightweight web framework. It also includes a shallow neural network
implemented with Keras for text matching and response selection. The application
processes the transcribed text, matches it against a database of possible responses, and
plays the corresponding pre-recorded video message.
The codebase already exists and will be provided upon hiring.
We are seeking a freelancer to implement key enhancements and prepare the app for
real-world iPad use.
The project is time-sensitive, so bids with a shorter timeline will be prioritized.
Key Tasks
1. Automatic End-of-Speech Detection:
- Remove the need for the user to press a button to end speech input.
- Implement an automatic end-of-speech detection mechanism.
- Ensure smooth integration with the existing backend flow that triggers transcription
and video response.
2. App Packaging for iPad Use:
- Build and package the Flutter app into an installable iOS format (IPA or
TestFlight-ready).
- Ensure compatibility and ease-of-use for caregivers and patients with no technical
expertise.
- Handle any permissions or audio-related configurations needed for deployment on
iPads.
- Provide simple installation instructions or an onboarding guide for caregivers.
3. Optional Improvements (if time and skills allow):
- Refactor the code and add extensive documentation for future developers expected to
continue working on this project.
- Provide packaging instructions and document any problems with their solutions
Required Skills
- Python (Back End):
• NLP experience, especially with NLTK.
• Familiarity with neural networks and DL tools (Keras).
• Experience with Flask..
- Flutter (Front End):
• Strong experience in building cross-platform apps.
• Integration of speech/audio functionality in Flutter.
• State management and asynchronous operations.
- Speech & Audio Processing:
• Understanding of end-of-speech detection or VAD (voice activity detection) techniques.
• Experience with audio stream handling in mobile environments.
- iOS Deployment:
• Experience building and signing Flutter apps for iOS.
• Familiarity with Apple provisioning profiles, certificates, TestFlight, or direct IPA creation.
Deliverables
- Updated well-documented codebase.
- Working app with automatic speech end detection.
- iOS installable package ready for testing and use on iPads.
- README and short installation guide.
Timeline (Estimated Hours)
• Automatic End-of-Speech Detection: 20–30 hours
• Packaging and Testing for iPad: 15–25 hours
• Documentation and guide preparation 3-5 hours
Total Core Task Estimate: 38–60 hours
How to Apply
Please share your relevant experience, particularly with: Flutter + iOS packaging,
Speech/audio processing, and Apps designed for health or accessibility.
We’re excited to work with someone who shares our passion for tech with purpose. We look
forward to receiving your proposal with the estimated timeline and expected budget. We
will schedule meetings with prospective developers to finalize the details.