The UCL MSP team recently ran a very successful event on Microsoft Cognitive Services: https://www.facebook.com/events/124243888288781/ (The event page has been viewed more than 20,000 times!)
GitHub repo for the the second part of the event can be found at : https://github.com/skabra07/MSP-TextSentiment
We would like to highlight the contributions of a number of MSPs at UCL to this event: Sidak Pasricha, Sondre Milch Agledahl, Kiran Gopinathan, Carmen Ibanescu, Suyash Kabra, Valentin Mihov, Sebastian-Valentin Burlacu, Rudraksh Shankaran, Dragos Fiera, Chris Obasi and Slaviana Pavlovich.
UCL MSP lead Slaviana Pavlovich had been looking to host an event at Canary Wharf, since the spring of the previous academic year. Thankfully, helpful members of the administration of UCL’s School of Management (Jemma Leahy, Susan Weatherley and Jackie Loveland) helped bring this into reality.
After careful planning, we decided on a suitable workshop topic. Custom Vision (https://www.customvision.ai/) is one of the more fascinating products provided by Azure Cognitive Services, and was one that fellow UCL MSP Sidak Pasricha happened to be knowledgeable about (read his blog post here: (https://blogs.msdn.microsoft.com/uk_faculty_connection/2017/12/01/utilizing-custom-vision-ai-for-creating-cognitive-services-projects/)).
From the beginning of our event, we wanted to make our attendees see the bigger picture, so we tried to make them think about artificial intelligence and whether this will change the whole software engineering industry. We then narrowed down to a smaller division of this topic, Azure Cognitive Services, more specifically, Custom Vision AI.
After showcasing one of our projects done at the MSP Summit using this framework, we wanted to let our attendees take upon a 50-minute challenge of developing their own classifier for whatever topic they wanted. The lucky winners were able to go home with the latest Raspberry Pis and a bunch of other swag provided by Microsoft.
After having a well-deserved pizza break, we concluded our event with a presentation regarding text analysis given by our fellow MSPs Kiran Gopinathan and Suyash Kabra. The program was developed in Python (https://github.com/skabra07/MSP-TextSentiment), and it was meant to show how AI can evaluate the text of an article and decide whether it is negative or positive, as in what is the general emotion one would get after reading that article.
We hope we provided a captivating introduction to artificial intelligence as a whole topic and hope that we pursued our attendees into wanting to learn more about this technology and its broad application to almost every industry out there.