Microsoft Data Platform Hackfest

Five customer teams came to the event with business problems they wanted to solve, or opportunities they wanted to pursue, using Azure data services and the Microsoft bot framework. Theta technical experts Jim Taylor, Sai Krishna, Umair Khan and Peter George, co-ordinated by Emmanuel Auffray, were on hand alongside Microsoft product specialists and technical evangelists to help the customers realise their goals and prototype solutions.

Here are some of the projects the Theta team worked on.

‍

Live sales data on a Power BI dashboard

Sai’s team built a Power BI dashboard that showcased live sales data using Azure Stream Analytics and historic data from an Azure Data Warehouse. She comments:

It was exciting to see an idea taking shape and being able to see a working solution within a working week. The data we were dealing with were quite big for usual NZ standards. Say 15 million transactions for one customer per week. The dashboard we built showed, up to the micro-second level, daily sales data and compared it to the previous day. We also explored the option of natural language querying which was quite exciting.

Computer vision and predictive analytics

Umair used his specialist skills to support two different teams. One was looking at predicting customer churn while the other was working on asset recognition using computer vision technologies. The teams used Microsoft Azure ML studio, R and Microsoft Cognitive Toolkit to complete their prototypes.

‍

Bots for operations – hands-free!

Jim Taylor has worked on several bots using the Microsoft bot framework over the past year, and used this experience to work on a bot that could help operations staff get information about jobs and tasks and schedule information. The team used new technologies, including the Microsoft bot framework, native browser implementations of text to speech and Microsoft’s speech to text service to make the bot work hands-free. Training language understanding models using Language Understanding Intelligence Service and SignalR real time communication was also part of this team’s work.

Solutions in just a week

The week ended with a show and tell event, where each team presented their business problem, showed a solution diagram and performed a live demo. It was impressive to see what could be achieved in just a week, with great technology and focused collaboration. Everyone took something from the event, from new business opportunities to increased awareness of technology and even prototypes for new solutions. Sai comments:

This entire experience was great opportunity to get to know real business problems and aspirations. It was quite fascinating to see what could be achieved by bringing business ideas, some tech-enthusiasts and good technology together. Overall, it was a great event to be part of. Everyone took away something useful from it.