Vision and wellness monitor mobile app for gamers

Research

Extensive research was conducted to address UI/UX challenges, incorporating user feedback to define features. A user-centric design approach ensured optimal solutions, guided by valuable insights gathered throughout the process.

Design System

Collaboratively established a robust and comprehensive design system optimized for subsequent changes. This in turn provides greater clarity and time-saving benefits for the development team with consistent components throughout the app.

Prototyping

Developed interactive prototypes used for user testing and validating assumptions. These prototypes empowered the client to gather valuable insights and facilitated efficient decision-making by stakeholders on feature optimization.

Cross-Platform Design

Implemented a cross-platform design approach that considered platform-specific UX requirements. The app seamlessly adapts to iOS and Android, delivering an optimized user experience while leveraging platform strengths for enhanced usability and engagement.

Proof of Concept Validation

Proof of Concept Validation: We built a Proof of Concept to ensure React Native’s suitability for this usecase. For the communication with the Bluetooth Low Energy (BLE) device, along using the already existing React Native open source packages, we also implemented custom native modules to fulfill the project’s needs.

Technology Stack

Designed the overall system architecture and carefully selected the technology stack, ensuring a robust and scalable foundation for the project’s success.

Feature Rich Mobile App

Developed the feature-rich mobile application, leveraging React Native to provide a user-friendly interface and creating iOS and Android native modules to achieve a seamless control over the Vivior wearable device.

Analytics

Displaying real-time metrics based on the data received from the BLE device.

Cloud Infrastructure Setup

Built a secure and scalable cloud infrastructure on Google Cloud, incorporating Firebase, Cloud Run, Dataflow, Cloud Functions, BigQuery and CloudSQL databases. It enabled data management, firmware releases, and efficient DevOps practices.

Extensive QA & Stress Testing

Conducted rigorous quality assurance testing to ensure the mobile app’s compatibility with the wearble device across iOS and Android devices. We also performed stress testing on the wearble device to make sure we have a reliable and stable user experience.

Agile Way of Working

Employing a SCRUM Agile methodology, we collaborated closely with the Vivior team, delivering work iteratively in two-week sprints. Regular reviews and demos ensured effective stakeholder management and alignment throughout the project.

Terraform-Based GCP Provisioning Security

The backend used Terraform to provision the GCP environments, which were separate projects and followed principle of least privilege.

Typescript API with Prisma

The Node.js API layer was written in Typescript, following Openapi practices and leveraging Prisma as an ORM to access the Cloud SQL database.

Commit Standardization and Code Integration

Alongside the Typescript code, commit messages were also validated and reviewed using commitlint and commit hooks and thus the entire team followed the same conventions and features were easily linked with JIRA issues.

Efficient CI/CD with Docker

The Bitbucket CI/CD pipeline built using Docker, tested and shipped the code in the cloud environment, ensuring code quality and predictable release patterns & times. Makefile was used for simplifying configurations.

Apache Beam ETL Integration

The Apache Beam (Dataflow) Python ETL pipeline ensured that the sensor binary data stored in Cloud Storage was transform and fed into Pub/Sub for other systems to consume, an example would be serverless functions, or stored into BigQuery for analytical purposes.

Comprehensive Testing Strategy

Dev testing was performed using Postman and also the dev team was involved in load testing using JMeter & Artillery, joining forces with the QA team to monitor the impact under heavy load and to apply performance fixes.

Joint Development

Due to the dynamic nature of this application, involving significant complexity for both its frontend and backend as well as an external sensor bank (wearable) with BLE connectivity and a proprietary algorithm which processes raw sensor data and returns easy-to-comprehend statistics, development has been multi-pronged and simultaneous for each level of functionality. As a result, the QA process mirrored the app’s development, involving Firmware tests for the external sensor bank, Functionality tests for its frontend, performance and stability tests for its backend and pipeline, as well as blackbox/greybox testing for the algorithm itself.

Connectivity

To provide accurate, real-time statistics that are easily digestible by end-users, robust communication between the external sensor bank, app, and algorithm is essential. Extensive connectivity and pipeline tests have been conducted to ensure stable communication among all components. This includes Load and Stress tests using Artillery and API tests with Postman and JMeter. Special attention was given to Bluetooth Low Energy connectivity, performing thorough connectivity, interference, and stability tests due to Bluetooth’s often unpredictable behavior.

Full data management and availability

the app’s intent is to not only process real-time sensor data and provide digestible reports for the end user, but to provide historic data as well, allowing users to track and monitor their progress on a day-by-day basis. Data reports, both real-time and historic, have been tested extensively in order to ensure they are always accurate and easily readable by anyone at a first glance.

Real life use cases

in addition to the standard testing process, which oftentimes involves brute-forcing functionalities and a significant amount of Edge and Corner Case validations, we’ve also tested the application within real-life scenarios in order to ensure that we’re not missing the forest from the trees. The entire team contributed to said tests, with everyone wearing the external sensor bank and monitoring our statistics while going about our daily routines. This was done in order to ensure that, no matter when and where, the application would always inform the user of their current status.

Continuous calibration

in order to refine the user experience and ensure that it is as accurate and as relevant as possible, the app has been constantly tweaked and polished – especially concerning data processing and delivery aspects, with the team offering many suggestions for improvements and changes which we felt would contribute to generating an awesome product.