FITTR Reduces Costs by 50%, Boosts Agility Using AWS

Founded in 2016, FITTR rapidly grew as a health and fitness platform. By 2022, the company recognized the need to modernize its infrastructure to support its growing user base and ambitious plans for expansion, including the launch of wearable technology, the HART ring. The company sought to enhance its ability to scale, deliver new features faster, and improve system visibility in preparation for future growth.

One of FITTR's first priorities was to address scalability. The company needed to ensure its platform could handle surges in traffic during marketing campaigns. In addition, with the planned launch of the FITTR HART wearable in 2024, the platform had to support continuous data streams and growing data volumes from these devices.

FITTR's monolithic architecture also limited its development agility, typically taking up to 4-6 weeks to launch new features. The accumulated technical debt from ‘dead code’ increased the risk of bugs, which slowed down development and testing processes.

Moreover, the company's previous infrastructure limited visibility into key metrics, such as error rates, response times, and system availability. As a result, the platform needed to improve its ability to resolve issues promptly and optimize resources effectively. As FITTR prepared to handle more personal health data from wearable devices, it prioritized enhancing system visibility and security to protect user data and meet evolving healthtech regulations.

FITTR initially built its app on a monolithic architecture but as the platform scaled, it found that it could no longer meet the demand for rapid feature development and increasing data loads. To resolve this, FITTR worked with AWS to implement a microservices-based infrastructure that could scale seamlessly and support growing user demands and data volumes.

The company used Amazon Elastic Container Service (Amazon ECS), a fully managed container orchestration service, to containerize its workloads and support scaling based on traffic demands. With this, FITTR could now handle traffic spikes of up to 1,000 times the average load during marketing campaigns. Additionally, the platform could manage continuous data streams from its FITTR Hart wearable, delivering seamless real-time data processing. By optimizing resource utilization, the platform's shift to Amazon ECS also helped reduce costs by 50 percent.

To reduce API latency and accelerate feature development, FITTR utilized Amazon API Gateway to create, publish, maintain, monitor, and secure APIs efficiently. With this fully managed service, the platform reduced the time needed to launch new features from 4-6 weeks to 2-3 weeks. FITTR also improved API performance, cutting latency from seconds to less than 150 milliseconds, creating a faster, more responsive experience for wearable and mobile users.

To optimize data management, FITTR deployed Amazon Aurora, a global-scale relational database service, and Amazon DynamoDB, a serverless, NoSQL, fully managed database. The company utilized both AWS services to handle the growing volumes of health data generated by wearables and mobile users. FITTR now had fast, reliable access to real-time personalized fitness insights. As a result, the platform could now deliver a seamless user experience across its app and wearable devices.

For real-time communication between microservices, FITTR adopted Amazon Managed Streaming for Apache Kafka (Amazon MSK), a fully managed, highly available Apache Kafka service. With Amazon MSK, the platform could now process data streams in real time. As a result of timely fitness insights, FITTR could deliver personalized recommendations to its users based on their latest health metrics.

The company used Amazon CloudWatch, a resource and application monitoring service, and AWS Security Hub, a cloud security posture management service, to improve visibility and security. With Amazon CloudWatch, FITTR's team could now proactively identify and resolve operational issues, improving system reliability. Additionally, by centralizing security alerts and automating security checks, the company met evolving healthtech regulations while safeguarding personal health data.

To give HART users insights to their health data and their internal users a view of the business metrics, Fittr also runs its analytics platform on AWS. They built a data lake using a combination of Amazon Simple Storage Service (Amazon S3) to help store, manage, analyze, and protect its data, AWS Glue, a serverless data integration service, and Amazon RedShift, to modernize their data analytics workloads and deliver insights. 

FITTR implemented its modernization strategy in phases over a year, maintaining business continuity and innovation throughout the transition. By phasing the migration, the company could test each workload in production, minimize user disruptions, and continuously introduce new features to its users.

With its modernized infrastructure, FITTR is advancing its use of generative AI with Amazon Bedrock to build a personalized coaching recommendation system. This AI-powered solution will analyze user data to segment users by their fitness goals and match them with coaches whose expertise aligns with their needs.

As the company scales this technology, it aims to revolutionize the user experience by delivering highly customized fitness journeys for each individual.

"The healthtech space in India is highly competitive. Staying ahead means continuously delivering innovation. With AWS, we've built a flexible, scalable foundation that not only accelerates our time-to-market but also positions us as leaders in shaping the future of fitness technology. AWS helps us to bring new products and services to life faster, setting the pace for innovation and pushing the boundaries of personalized health and fitness experiences,” said Mandar Nilange, head of engineering at FITTR.