By Anahita Bhiwandiwalla, Sr. Software Engineer
The Data Science Summit was held on July 12th and 13th in San Francisco with around 1,400 attendees. It was an incredibly vibrant atmosphere being around 900 data scientists and developers in addition to business executives. The platinum sponsors of this event were Turi, Intel and TAP. Like other technical conferences, the sessions were divided into Advances in ML Research, Data Science Use Cases, Systems for Data Engineering and Tutorial Sessions spread across two days.
The event also had talks from leading researchers in the field of Machine Learning like Prof. Mike Jordan from UC Berkeley and Prof. Carlos Guestrin from University of Washington to name a few. There were also industry experts like Jeff Dean from Google and Dr. Xavier Amatriain from Quora and others.
Kyle Ambert, a Data Scientist from the TAP team, gave a 20 minute talk on “Introducing the TAP Analytics Toolkit”. Kyle’s talk had a packed room full of attendees eager to understand what TAP is all about. Dr. Iman Saleh, a Big Data Evangelist for TAP, and I conducted a 1.5 hour training on TAP, walking everyone through an anomaly detection use case. The session had around 35 attendees and it was very well accepted. We gave a training on the different components of TAP, how anomaly detection can be performed on real world IOT sensor data using TAP, how applications can be deployed on TAP, and what different data science problems can be solved on TAP. We covered a lot!
Robert Scoble, one of the most known technology bloggers came to this event to learn about TAP. He interviewed me and Kyle on TAP and Data Science. The interview is available on Facebook. The talk has got 3900 views in just one week!
We gave demos of TAP at our booth that covered an overview of the various components of TAP. Our booth at the conference had our Marketing team along with an Engineer and Data Scientist at all times to help cater to the broad audience. So next time you see us at a conference, do not be afraid to come ask us a question or two.
What impressed me most about the event was the breadth of talks ranging from inventions in academia to novel usages of data sciences to how to actually get it done; all provided an entire workflow of data science knowledge, no matter where you fit into the pipeline. Some of the talks that I attended spoke about how to help the adoption of data science, how to build trusted between users and the system, and how to personalize predictions and recommendations. Once users begin to trust the predictions and understand what is influencing a decision, they are more likely to adopt it compared to a black box making evaluations. I also really liked the Startup Sessions they had in the latter part of the day which had 10 minute lightning talks by upcoming startups addressing the problems they solve and how they do it.
I feel like the industry is increasingly moving towards more intelligent applications and personalized predictions – ranging from predictions about one’s health, to continuous predictions and recommendations from an autonomous vehicle, to playing one’s favorite music on your internet radio.
A platform like TAP makes it easier to do that:
Reduces the overhead of dealing with compatibility issues in services that your application needs
Very convenient to access a wide range of services
Easy to reuse feature engineering methods a data scientist may develop
Convenient model deployment and scoring (plus more!)
While conversing with data scientists at the summit they stated that these value adds that TAP provides would really enable them to focus on applications to be developed and increase the speed of building innovative solutions which is exactly what we need them to do.
Checkout my bio from the conference.
Join our community at https://community.trustedanalytics.org and be sure to watch our event calendar at http://trustedanalytics.org/events/ to keep up with our latest meetups, webinars, and conference events.