From Connected Health Data to Personalized Patient Intelligence
Tizora engineered a digital weight management platform for a gastric balloon program that syncs Bluetooth scales automatically and unifies weight, sleep and exercise data in one patient view. Patients get progress reports and in-app chat and video with healthcare professionals, and weigh-ins reach their record in under a minute.
What the platform changed for patients and care teams
What problem was the gastric balloon program facing?
A gastric balloon program does not end with the procedure. The patient's success depends heavily on what happens afterward—whether they follow recommended lifestyle changes, remain physically active, monitor their progress, and continue engaging with healthcare professionals.
The traditional approach creates several problems: manual weight tracking, distributed activity data, limited visibility for healthcare professionals, and fragmented communication.
The product needed to solve a bigger problem than simply "track weight." It needed to create a continuous digital feedback loop between patient data, patient behavior, and healthcare guidance.
- Weight tracked manually by patients
- Activity, sleep and exercise data spread across different sources
- Limited visibility of patient progress for healthcare professionals
- Fragmented communication between patients and experts
How did Tizora build it?
We engineered the ecosystem behind the screens: collecting data automatically from connected devices, normalizing it into one patient record, turning it into insight and connecting patients with healthcare professionals.
How is the platform architected?
Cloud-based backend services run on AWS, AWS Lambda, Python and PostgreSQL. Core capabilities are separated into services that can evolve independently, and an integration and normalization layer turns data from Bluetooth scales and other health sources into one patient data model shared by the iOS and Android apps, progress reports and chat and video consultations.
The architecture leaves room for an intelligence layer above the data platform. Instead of delivering identical content to everyone, an AI personalization engine can determine what a patient should see next based on historical data, trends, patterns and risk signals, turning scheduled communication into context-aware engagement. It is the same idea behind our AI skin analysis feature, which matches each shopper to the right products.
The complexity of this project was not in building screens. It was in engineering the ecosystem behind those screens. The same engineering-first approach shaped our AI-ready patient management platform.
What does the platform do?
We transformed raw health data into visual reports, progress patterns, and a connected care loop.
What technology powers the platform?
Mobile & web
Cross-platform patient experiences across iOS and Android.
iOS
Android
React
Building a connected health or remote monitoring product? See how this architecture fits your devices and care workflows.
Outcome: automatic weigh-ins and one view of patient progress
The platform replaced manual weight tracking and scattered health data with automatic device sync, one patient data model and in-app communication. The figures below are indicative operational estimates, not clinical outcomes.
| Metric | Before | to | After | Change |
|---|---|---|---|---|
| Weigh-ins entered by hand | 100% | Under 5% | −95 pts | |
| Time for a weigh-in to reach the patient record | Next manual entry | Under 1 minute | Near real time | |
| Places a patient checks for health data | 3 or more | 1 | One view | |
| Care-team time to review a patient's progress | ~10 min | ~4 min | −60% | |
| Remote consultations | Outside the app | In-app chat and video | Integrated |
How we measured
- Indicative figures based on project estimates; to be replaced with measured client data.
We engineered the ecosystem behind the screens: Bluetooth scale sync, one patient data model for weight, sleep and exercise, and in-app chat and video, so patients and healthcare professionals work from the same picture of progress.
Lessons
What we learned
Normalize before you analyze
Devices and APIs report in different formats and frequencies, so the normalization layer had to come before any progress reporting.
Remove the manual step
Syncing the scale over Bluetooth removed the manual weight entry that the old tracking approach depended on.
Structure data now for AI later
A unified data model and service-oriented architecture let personalization and risk scoring be added without a rebuild.
Will this work for your digital health program?
Who this approach fits and what it connects to.
- Built for
- gastric balloon and similar programs that need continuous monitoring, behavioral support and expert guidance
- Integrates with
- Bluetooth scales, activity, sleep and exercise data sources, and Twilio chat and video
- Typical scope
- iOS and Android patient apps, AWS cloud services and progress reporting for healthcare professionals
This approach fits weight-loss and lifestyle programs where success depends on monitoring and supporting patients between visits.
Digital weight management FAQs
Frequently asked questions
It is software that brings a patient's weight, activity, sleep and exercise data, progress reports and contact with healthcare professionals into one app. In this project it supported gastric balloon patients after the procedure, when success depends on lifestyle changes and continued engagement.
Patients weigh themselves on a Bluetooth-enabled scale and the reading syncs automatically. This removes manual weight entry and creates a continuous stream of weight information.
An integration and normalization layer converts data with different formats, timestamps, frequencies and measurement patterns into one unified patient data model covering weight, activity, sleep and exercise.
Real-time chat and video consultations built with Twilio run inside the app, so remote expert support is part of the patient's digital journey.
Cloud backend services on AWS and AWS Lambda, Python microservices and PostgreSQL, React, patient apps for iOS and Android, and Twilio for chat and video.
Yes. Structured data and a service-oriented architecture provide the foundation for an AI personalization engine that decides what a patient sees next based on historical data, trends, patterns and risk signals, and for capabilities such as risk scoring.
Disclosure
Client name withheld under a confidentiality agreement. Results reflect this client's data and will vary. Some figures are indicative estimates pending measured client data. This case study describes software engineering work and is not medical advice.
More healthcare case studies
Building a connected health or remote monitoring product?
Tell us about your devices, data and care workflows, and we will suggest an architecture and a phased build plan for your platform.
