eWeekchat

James Maguire

And then there's the issue of quality of data...

challenges de-duplicating and harmonizing data sets from disparate applications creates big #dataquality issues

Andi Mann

@JamesMaguire So true. Never is 'Garbage In, Garbage Out' more applicable - it is exponential at the scale and speed of #analytics !!

Tapan Patel3

A3 I would biggest pain point(s) are not even related to technology! They are related to poor data literacy, culture challenges to accept change, skills/staff shortage. #eweekchat #DataScience #analytics #AI #datamanagement

Andi Mann

Aha!! Always the biggest issues. Technology is easy - people are hard!! 😁

Rex Ahlstrom3

A1. Data Fabrics are becoming a priority. This will help cross application analytics access more data in a faster and simpler way.

James Maguire

I hear a lot about data fabrics. Almost a "topic du jour"

A lot of hype around #DataFrabric but also a lot of potential if implemented properly.

A #DataFabric is also not something you just "buy and install". It is a design pattern that needs to be considered alongside existing systems & technologies.

Tapan Patel2

A5 - Myth: AI is all about deep learning. Reality - Majority of problems are still solved using traditional ML methods and techniques.
#AI #analytics #datamanagement #datascience #artificialintelligence #eweekchat

James Maguire2

Thanks for bearing with us.

Bruce Kornfeld

A5 Data analytics eliminates human bias - depends on how and how it was set up

(edited)

Andi Mann

@brucekornfeld LOL, mostly exactly the opposite - #analytics frequently reinforces human bias!

Tapan Patel2

A4 - Data Science teams should not be building models or working in isolation. Collaborate with your peers in data management (i.e. data engineers) and IT (e.g. ITOps) for model deployment. #AI #analytics #datamanagement #datascience #artificialintelligence #eweekchat

Tapan Patel2

A3 Lack of involvement, support and ownership from business stakeholders is another major barrier. #eweekchat #DataScience #analytics #AI #datamanagement

"I hired a data scientist, why don't I get good analytics results"... executive sponsorship, #dataengineers, #dataanalysts, data-driven culture... spot on

Rex Ahlstrom2

A3. @Chris_Ehrlich I completely agree. Simple access to the high-quality data they need while keeping an eye on #governance and #compliance.

Rex Ahlstrom2

A2. Many #analytics programs are delivering value but falling short. Still too siloed, departmentally focused, limited by access to high quality data. Proliferation of compliance issues also making it much harder to get the job done and stay out of the courts.

@Syniti is constantly being asked by customers how to make it easier to source high quality data for #analytics, avoiding the garbage in garbage out challenge.

Andi Mann2

A2. Think most are comfortable w/ elementary analytics. e.g. in #video, #performance is easy to analyze, take action on insights to improve #UX