Case Study: CloudEngage scales on AWS
CloudEngage is a Wi-Fi Engagement platform with Wi-Fi and BLE IoT Analytics capabilities.Tangent Solutions resolved several challenges facing the CloudEngage team, enabling them to focus on perfecting their data processing and analytics craft.
The Challenge
CloudEngage believes in unlocking the potential of statistical analytics, specifically in the domain of in-building location data.
Its goal was to release a single solution for receiving and processing wireless beaconing data from 1000’s of Cisco Meraki wireless devices.
The platform also has a number of Authentication / Authorisation components including Splash Page and RADIUS. One of the unique features of CloudEngage it its ability to authentifcate users onto WPA2-Enterprise Wi-Fi using their existing Office365 credentials, without the need to have any form of AD.
There are also public Wi-Fi capavilities; when users attempt to connect to a Public Wi-Fi SSID they are redirected to a Splash page where they can register/login, acquire a voucher, purchase access, etc. Additionally, network users can see advertisements and complete surveys before being granted access to the internet.
Multiple Infrastructure Providers can support multiple Service Providers’ SSIDs on their hardware, further allowing them to run many concurrent campaigns and splash portals.
Each of these permutations of SSID, Infrastrugture Provider and Service Provider and the concepts of Tags, makes for an extremely complex data processing requirement.
Unfortunately, they faced constraints in the form of performance and infrastructure limitations.
Data capturing and processing often requires event-driven and dynamically scalable solutions. Traditionally, CloudEngage would need to provision infrastructure and architect for a maximum load to avoid performance drops.
Re-architecting the solution to leverage the inherent scalability of the cloud was able to solve the scaling problems. The lack of elasticity limited their initial solution in terms of expanding to larger networks, and allowing better coverage.
The CloudEngage team had enough on their plates to deal with interesting ways of treating the data, and building continuously learning ML Models for improving location accuracy. They found relief in the fact that they could leverage the AWS expertise of Tanget to handle the Cloud Architecture and Scaling, while they focused on their core offering.
The Solution
AWS was an obvious fit for the server management and scaling needs of CloudEngage. With EC2 and Auto Scaling, a fully managed service was delivered, allowing the client to have the confidence that any performance load would be automatically catered for.
The cloud resources were further configured to be highly available, fault tolerant and durable. This was also an improvement of their original design, which did not have HA natively designed into the solution.
The solution easily extended to and integrated with AWS powered databases, and an API. AWS serverless compute, Lambda, allowed for a decoupled, event-driven interface between the API and the internal services.
Fleets of RADIUS servers on EC2 enabled user authentication and accounting. Various features were included in the solution focusing on user customizations and “white-labelling”. Most of the analytics data was stored in AWS RDS and is interacted with using Lambda.
With a strong dependency on IoT and Big Data, CloudEngage was designed to plug in seamlessly with other services using AWS API Gateway.
AWS Services Used
AWS Architecture
The Outcome
CloudEngage initially dedicated its development capacity towards analytics and processing functionality. With a shortage of staff and lack of time, automating AWS deemed to be too complex for immediate use. Tangent Solutions filled the void with their experienced and certified experts, while greatly improving AWS adoption within the CloudEngage team.
The move to AWS allowed CloudEngage to specialise in what they are comfortable with, and the rest was taken care of. CloudEngage could now focus their efforts in selling their product with trust that is backed by powerful infrastructure.
