Enabling Customer Driven Application
Development for Gett B2B
Overview
After carrying out a joint market study, Sphere developed a successful customer-oriented prototype for Gett Taxi.
The Challenge
Gett Taxi wanted to explore a deeper integration of their technical ecosystem that addressed the needs of corporate clients, especially key strategic clients. Specifically they sought to incorporate these use cases for an improved internal platform:
Advanced BI and Analytics
Easy Pricing/Tariff Management
Corporate Payment Methods
Enabling seamless customer support
How It Was Solved
The Prototype was at it's core, designed to be very flexible. It addressed the needs of the corporate clients and additionally provided the capability to add more functionality using the system and user settings. Sphere designed a modular architecture for the application which allowed for any of the key components to be extended by its users or partners.
- For the internal platform entities and external interactions with the payment gateways, vendors and other third-party services, Sphere implemented varying levels of logging using Ruby gem (audited type), time-series storage ClickHouse and some DBMS Postgres capabilities
- To provide optimum performance during uneven load increases while running multiple background tasks, Sphere used: AWS metrics and integration with NewRelic
- The configuration of the puma web server was optimized, as well as the grouping of background processes in Sidekiq
- Separate processes and queues were selected, limits were set on the parallel execution of tasks and consumed resources
The Results
Four primary benefits were created from the functional prototype including:
Self-sufficiency and high levels of customer satisfaction
Gett Taxi now integrates new strategic clients quickly and seamlessly into their platform which is customised to clients business processes and workflows
Performance
Sphere set up a well-established process for analyzing incidents related to performance:
1) Identifying problem areas, optimizing and allows for support teams to promptly deliver fixes to the production 2) It also demonstrates 99.9% uptime, average server response time is approximately 200ms
Scalability
The application architecture is aligned to the business needs and forecast thus giving Gett Taxi the capability to extensively scale both vertically and horizontally
Log analysis
A comprehensive logging set of all users and system actions: Customer service teams can track any order status or other entities, providing fast resolution for an disputes with the suppliers of the external services
