Improving Experiences of Micromobility Customers Using an Advanced Data Platform

More Advanced Cloud Infrastructure, Faster Data Processing Capabilities for Real Micromobility

Swiftmile, a universal charging platform for micromobility, was preparing for the next phase of its worldwide expansion. The company was evaluating new market opportunities and testing different business models in the dynamic market environment. The leadership was also exploring partnership opportunities with such global brands as Spin, Lime, and Bird.

However, Swiftmile’s potential for growth was limited by its own micromobility platform. The team realized that the platform’s backend logic and monolith infrastructure had to be overhauled, and the platform itself had to be substantially modernized. They needed to design and build a robust, highly scalable, cloud-based platform that would enable them to:

• Maximize scaling potential, to cope with any demands for business growth, including adding new stations and third-party integrations
• Simplify the integration of new incoming and outgoing communication streams, such as OEM systems protocols, Salesforce, etc.
• Prepare for worldwide expansion, both technology- and business-wise, while keeping down the costs of handling the platform’s infrastructure

Aside from improvements to the backend logic and infrastructure, the team at Swiftmile sought to make their platform data-driven. Specifically, they wanted to gain the ability to collect, process, and analyze streaming data in real time, to provide higher-quality service and improve user experience. In addition, the improvements would help the company to explore data business opportunities by making it easier to add open data sources and integrate routing features of Google Maps, to ensure genuine micromobility.

To support the existing scaling commitments defined in contracts and technical needs of new business verticals, Swiftmile reached out to Provectus, an AWS Premier Consulting Partner with Data & Analytics Competency. The teams started with a discovery workshop to assess Swiftmile’s current state of business and technology, and to define a comprehensive strategy to move forward.

Designing and Building Swiftmile’s Micromobility Platform with Data Streaming Capabilities

The initial assessment of Swiftmile’s micromobility platform consisted of a comprehensive review of the system environment, data infrastructure, data flows, and the logic of communications processing. It involved running a series of load tests, to help define the maximum number of stations and maximum messages per second that the platform could handle. Various testing scenarios were employed to identify main bottlenecks of the platform.

It was discovered that the monolithic platform could not support more than 600 stations while handling about 30 messages per second. Though the database was stable at peak load, the logic of processing of backend messages and the infrastructure itself were causing substantial time losses when processing transactions beyond 30 messages per second. This major bottleneck limited the platform’s performance, scalability, and cost-efficiency.

The Provectus team came up with the following solution:

1. Reduce the number of transactions and database operations in the message processing pipelines. The idea was to split the logic for handling stations and related messages into two independent parts, where each part is responsible for a single operation.
2. Transform the backend logic and infrastructure for data streaming tasks, using the best data streaming and data processing solutions.
3. Prepare the infrastructure to integrate advanced solutions for handling autoscaling and geoscaling components of the platform.
4. Use Provectus R&D products to reduce development and maintenance costs, minimize technical risks, and accelerate software release cycles.
5. Take advantage of best practices and industry standards for security to prepare the platform for possible security reviews and audits by partners. Make the platform ready to securely store third-party and private data.
6. Upgrade incoming and outgoing data streaming interfaces for integrating new data sources, including data from third parties.
7. Improve fault tolerance of the platform, to ensure that it continues to operate properly in the event of a failure of one of its components.

To realize these improvements, Provectus recommended using a toolset of AWS services, third-party tools (Apache Kafka, Java stack, etc.), and Provectus products, including Swiss Army Kube (SAK) and NextGen Data Platform.

The new platform was designed and built as a microservice-enabled, cloud-based application with advanced data streaming capabilities. The data streaming infrastructure was developed using best practices for storing, collecting, and managing streaming data on AWS. The platform and its new features were deployed, migrated, and implemented on production servers. Old production traffic was safely and securely switched to the new AWS environment.

The teams agreed that Provectus would continue to provide urgent product and technical consulting services for the newly developed data platform.

Modernized, Data-Powered Micromobility Platform Enables Faster Business Growth in New Markets

Provectus finalized the major portion of the platform modernization work within seven months of the project’s inception, to enable Swiftmile to start scaling their business as quickly as possible. All preliminary hypotheses and concepts of the team’s experiment were proven, and all business goals were achieved.

Swiftmile’s modernized micromobility management platform received advanced capabilities for handling streaming data in real time. It now can also collect, process, and utilize various types of data for integration.

The infrastructure of the platform was redesigned and rebuilt from scratch. It enabled Swiftmile to increase the platform’s performance and capacity up to 20x while retaining cloud infrastructure costs. To demonstrate its potential, the Provectus team also ran a series of experiments, proving that the platform could handle over 10,000 connected stations. 

The platform is now ready for almost endless cost-efficient scalability and load growth. Swiftmile can quickly scale business use cases and models, develop new functionality, integrate data sources, and add pipelines without having to change the overall structure of the platform. This also means that the company can easily scale business operations geographically — in Europe, the Middle East, and North America — simply by rolling out platform clusters in local regions.

There were also improvements from a customer experience perspective. The upload speed of station firmware has increased by 4x. The data extraction and processing speed has risen by 5x. Basically, this means better UX for end users and system operators, and reduction in operational costs with better fault resistance.

From a development perspective, the new platform has helped to noticeably reduce development and delivery overhead, minimize operational costs at all levels, and maintain zero-incidence performance, with no detected data losses and no incidents detected by the Provectus NOC team. The platform demonstrated 100% system availability.

Overall, Provectus’ work on the modernization of Swiftmile’s micromobility management platform has enabled Swiftmile to initiate and realize a business turnaround strategy. The new platform added scalability to ensure faster, more efficient growth; optimized operational costs (for both development and business teams); and improved customer experience through rapid access to data and its business-critical insights. Now, Swiftmile is fully ready to scale its business globally while exploring opportunities to transform into a data-oriented company.

Moving Forward

1. Learn more about the Provectus NextGen Data Platform
2. Explore more customer success stories covering platform modernization: IMVU, Blue Bottle Coffee, Model N, TripActions
3. Apply for NextGen Data Platform Acceleration Program to get started