IMML at Grace Hopper, Houston

  • Article

This post is authored by Rama Raman, Senior Software Engineer, and Rhonda Phillips, Program Manager, at Microsoft.

The Microsoft Information Management & Machine Learning (IMML) team had a strong presence at the Grace Hopper Conference, held recently in Houston. Here’s a short post highlighting some of our team’s activities there.

Booth & Presentations

Monica Kei, Software Engineer in IMML, and Rama presented some of the latest Cortana Analytics technologies and demos at the Microsoft booth. Our talk sessions were well attended and our idea of printing out customized How-Old.net photo cards for attendees created just the right buzz before our presentations!

Rama shared a story based on a presentation that had been delivered by Joseph Sirosh at his Strata keynote. The audience was intrigued, more people started to gather around our booth and they seemed to enjoy the How-Old.net story. They also realized how easy it is to build out their scenarios using the Azure ML APIs – Rama demonstrated how to integrate Face, Speech and Recommendation APIs to perform sentiment analysis using a [simulated] Intelligent Mall kiosk. Net connectivity at the booth, however, turned out to be spotty :-( and that forced us to use a recorded version of the demo – nevertheless attendees loved what we showed them.

Monica followed up with a brief description of Cortana Analytics Suite and how the various technologies work together. She also did a great walkthrough of our Connected Cars demo, showing how Azure Stream Analytics processes real time car data which can then be used by drivers to deal with issues on the road, or by dealerships to pre-order parts, etc. Meanwhile, big data in Azure blobs orchestrated by Azure Data Factory are used by insurance companies to determine insurance premiums based on the patterns that they see. Attendees grokked the power of these technologies coming together to solve real-world IoT scenarios such as these.

After our presentations, we had many attendees inquire about Cortana Analytics, with several of them expressing an interest to join our team! New graduates seem to especially love the data space and several were eager to be a part of the action! We are hoping to interview and hire some of the really talented people we met during the course of the event.

Fun with Machine Learning

Danielle Dean, Senior Data Scientist Lead in IMML, and Rhonda were invited by Jennifer Marsman, Principal Tech Evangelist in Microsoft’s Developer Experience Group, to participate in the Fun with Machine Learning ‘node’ that Jennifer and team had put up. Nodes were a new concept at Grace Hopper this year – the intent is to create an environment where people can gather and learn about and discuss a topic (such as ML) together. Our node was a big hit and we were at full capacity during our sessions.

At the node, we combined exciting stories about ML and what it can do along with mini lectures and hands-on ML exercises. Students and professionals alike were excited about the application of ML to a wide variety of scenarios, including:

  • Lie detection: Jennifer shared her work on lie detection using a super-cool EEG headset!
  • Dairy farming: Jennifer shared Joseph’s Strata keynote referred to above, on connected cows.
  • Hiring from a diverse application pool: Charna Parkey from textio showed how she uses ML to help companies craft job ads that are likely to increase the diversity of the applicant pool.

While learning about these awesome and important applications – and the one about increasing diversity really struck a chord with this audience – we also learned the fundamental concepts of ML algorithms:

Lauren Tran, Tech Evangelist at Microsoft, and Jennifer gave us an overview of supervised learning concepts and algorithms, while Rhonda talked about unsupervised learning and clustering. Dr. Sheila Tejada from the University of Southern California showed us how information theory can be used to build decision trees – and we then built our own version of these to address problems such as how to classify words and whether or not to wait for a table at a restaurant.

Fun with Machine Learning was an overwhelming success. We received great feedback, and – more importantly – the most popular question at the end of the session was “How do I learn more about machine learning and become a data scientist?” Our node was in such high demand that we’ll definitely need to plan for a few more sessions and more space at GHC next year (-:

Rama & Rhonda