Somehow Microsoft’s Build 2017 Conference has already come and gone and my head is overflowing with all the actionable insights, the creative discussions and the memorable presentations from some of the smartest developers in the world.  Big things were unveiled and promoted at this conference and not only did I get a chance to witness it, I also happened to be one of the lucky ones to present on my experience this past year with Sierra Systems and Pure Technologies.

That’s right.

Just a few weeks ago, I announced that I was selected to present at Microsoft Build Conference and did so just a few weeks ago; if you haven’t had a chance to read my previous blog post, What is your excuse for not going to Microsoft Build, the developer event of the year, be sure to check it out.

This was a once-in-a-lifetime opportunity, and to say that I wasn’t the least bit nervous would be a lie! I had to weave a 6-month project into a smooth 15-minute story – complete with detailed insight into the technical approaches that I encountered as well as a start-to-finish implementation demo of solution code. Fortunately for me, I had a wealth of experience and an engaging topic to deliver on.

The driving theme of the conference was, empower the world to achieve more. As developers in today’s world, we have a greater accountability than ever before for the algorithms and experiences we create to solve problems. The work we do has an impact on the companies we represent, the organizations we add value to and the world in which we live. In my presentation, I highlighted how my team used technology to make an impact – to change the world.  So, without further ado, keep reading to learn how we did just that. 

Data Analytics and Machine Learning in Action:

The subject of my presentation was data analytics and machine learning and how we applied it at Pure Technologies.

The Challenge:

The challenge that Sierra Systems was presented with was to minimize the classification time and effort required for technicians to identify oil and gas pipeline leaks.

The Solution:

Through data analytics and open-source machine learning, we developed a deep-learning artificial intelligence system using Python and the Microsoft Cognitive Toolkit. Our system converted the audio gathered by Pure Technologies sensor ball into visual representations which accurately identified leaks more than fifty times faster than a person could, and can be scaled to process multiple pipelines in parallel. With this new system in place, infrastructure managers could execute pipeline repair orders within hours as opposed to weeks.

Final Thoughts:

In summary, the past several months leading up to this defining moment have opened my eyes to what we can accomplish with a bit of creativity, passion and a lot of collaboration.  I wanted to send out special thanks to everyone who took the time to drop by our presentation, to my team at Sierra Systems and those that worked with me to bring about the Pure Technologies project and Build 2017 presentation to great success.