Rocket.Build 2020: Highlighting our Finalists Part 1
Although Rocket.Build 2020 presented its own set of challenges—building and collaborating remotely, dealing with unprecedented real-world challenges, and more—it also managed to bring new successes that will carry on to future years. With the event being entirely virtual, we were able to create global teams and have more participants than ever. Overall, 450 Rocketeers from 14 countries came together to build projects that represented 50% of all functions at Rocket. And we call that a success!
With the final round taking place this week, we wanted to highlight some of the 51 fantastic projects that have made it to round two. These projects power legendary results, create more secure environments and make our customer’s lives better. Without giving too much away, I’m proud to share them with you.
From MV to Kafka with Love
Team members: Vinnie Smith, Jenny Yao, Jianfeng Mao, Henry Unger, Stuart Shepherd and Adam Astle
Most Rocket.Build projects are built by a team of Rocketeers. This project was unique in that it involved two Rocket partners on the team—Stuart Shepherd and Adam Astle of APT Solutions and Henry Unger, CEO of Pulsiam—alongside three Rocketeers. Pulsiam builds software solutions for emergency responders, allowing for greater communication between responders and their units, which is critical to saving lives. APT has a solution that supports membership solutions, including some of the largest union bodies, which can help people stay connected especially during these pandemic times.
With their Rocket.Build project, they wanted to easily integrate real-time data with 3rd party data models. Using Kafka as an open source distributed event streaming platform, they worked with the two MV customers to see if they could stream topics to each other’s application that the subscriber could then ingest into their application. As a proof of concept, they integrated APTs membership system with Pulsiam’s Public Safety application. New members in the APT solution would appear in Pulsiam’s system as they had subscribed to Member data events. The goal of the project for APT is to allow for increased integration by reducing costs and by sharing or consuming 3rd party data easily from other applications such as HubSpot. This can help to increase marketing and other sales-related activities. Longer-term, this could also act as a generic transport for Public Safety Systems by Pulsiam, especially in the growing private medical response market where disparate tools exist across those providers and they need a unified data interchange platform. One main benefit of the Rocket.Build project is that no coding changes are required and MV customers can use any data streaming platform via Kafka to leverage existing MultiValue data to new workloads.
BlueZone Multi-Factor Authentication
Team members: Mike Jones, Mike Slater, Zhi Li and ShaoGeng Yi
When using Rocket/IBM multi-factor authentication with a terminal emulator, the workflow is awkward. It requires a user to open up a URL in a third party browser, successfully answer all the security prompts and then manually copy/paste the token from the browser into the correct field in the terminal session. With MFA becoming increasingly important with remote work becoming the norm, this team wanted to add capabilities to BlueZone that would simplify the MFA login experience.
In BlueZone, they wanted to add capabilities to detect when a user was entering the login screen to initiate MFA, without needing to switch to an outside browser. They also wanted to allow users to re-use the token when possible, without keeping the token in the Windows clipboard.
AI-Driven Prediction of Service Disruptions in Infrastructure
Team members: Abhishek Lokhande, Lalit Bharate, Suyog Jadhav and Parag Bhayani
Cloud Management Console (CMC) runs as a IBM cloud-hosted service to monitor complete Power Systems infrastructure. It provides dynamic views of performance, inventory and logging, all in a single location.
This Rocket.Build team wanted to introduce iAlert (Intelligent Alerts), an AI powered notification service (email, SMS, in-app), to predict infrastructure disruptions. It enables CMC users to take preventive actions, instead of being reactive to disruptions. It would also provide the opportunity for better planning of upgrades and deployments. In addition, dynamic threshold values calculated using anomaly detection techniques will reduce the requirement for manually setting up threshold values for thousands of Servers and VMs. iAlerts were tried out with several supervised machine learning algorithms. The team picked the “Random Forest” algorithm due to its high performance. They plan to improve its accuracy by combining it with back propagation (BP) neural networks and are excited to see its positive impact on CMC customers.