#DataOps - CrowdChat

DataOps in Action

Join DataOps experts and the IBM Chief Data Office as they share best practices and methodologies to deliver business-ready data fast to accelerate digital transformation and the journey to AI.

#dataops @dvellante @aliye1ozcan @karenmadera44 @vstasiewicz @chrisbergh @brianhforr @stevelueck @valuefromdata @elainehanley @bigdata_paulz @julielockner @tumimonale @dmenningervr @451research

LeaderBoard

#DataOpsDevOps to DataOps DataOps can be the next big Team Sport, lets build a dialog for DataOps

Stream Ended

Establishing a secure connection

Paul Zikopoulos39

#DataOps is a requirement to get the ROI on your data science teams ... you'll never get the ROI of a data scientists' high salary unless you take away the time they spend on DataOps & put it into a focussed service: data science is a team sport & this team needs a DataOps player

[+] Show Hidden Comments

A key aspect of #DataOps is instrumentation - which allows ROI to be effectively measured

(edited)

#DataOps optimizes #DataScientist skills andtime

totally agree - wrangling data gets old - even for data hackers...

@julielockner Our research is suggesting that data scientists are spending nearly half their time just finding and preparing data. If you can streamline that process via #dataops, you can immensely boost productivity.

Paul Zikopoulos

Yup! I need my data scientists search the hyper parameter space for better model performance; give them a shop for data approach with lineage and service-based flows that do the work for me.

Bradley Shimmin

It seems that what we're asking for here is perhaps something akin to expanding AutoML to encapsulate data prep.

DataOps is more than just for Data Scientists -- Data Engineers, Analysts, BU, Governance or anyone in the data value chain

Jay Limburn 🇪🇺

An intelligent catalog that provides that self service, automates the governance and turns the output of the governance initiative into insight for the data scientists is at the heart of a good #dataops strategy.

David Menninger

@julielockner ditto. Our research shows accessing and preparing data is the most common challenge faced in applying #AI/#ML https://www.crowdchat.net/s/260a0

Paul Zikopoulos

So there is surely a part of this that can connect - AutoML will look at the data and suggest some algorithms, but also great AutoML platforms inspect the data, report on it, fix it (suggest), identify, record metadata, +++

Dave Vellante34

Q5. How are you balancing business goals with compliance measures? Do compliance edicts restrict your ability to achieve your goals?
https://www.crowdchat.net/s/060es

[+] Show Hidden Comments

David Menninger

More than a third of participants in our research (36%) cite compliance as one of the reasons for changing their data processes

(edited)

With regulations like #CCPA, #DataOps helps make compliance affordable and possible

#DataOps done right makes governance an accelerator and not an inhibitor or roadblock...probably THE most important reason to implement #DataOps

(edited)

Policy management is a key step in the #DataOps process. A smart catalog knows who should be able to access the data to keep the right side of regulations

I think it's a misconception that compliance needs to be "balanced" with business goals, as if they are opposing forces. Good governance and #DataOps methodology helps achieve compliance and help the business move faster.

Requlatory compliance with data privacy laws is one of the key drivers for Data Ops --- controlling where data is located and who has access ------ rules and policy integrated with authentication and controls

@dmenningervr can they turn those compliance requirements into other business value Dave? Or is that a pipedream?

Compliance can be a reason to start - but real business value is what differentiates

Just the contrary. I recall @TumiMonale said a compliance crises helped them get their data foundation implemented.

Absolutely. A lot of compliance objectives align with other needs, such as data quality. A DSAR looks a lot like a customer 360 view.

Compliance and Data Democratization seems to be at odds but story is a bit inflated most of the time. There are tools out there which can help us anonymize the data

Paul Zikopoulos

@SDobrin It's nothing less than shocking how many times I see clients waste an opportunity because they are just trying to check the compliance box - they do the hard work and lose the big time benefit beyond the obvious.

(edited)

David Menninger

Not mutually exclusive. Our research also shows that better data processes lead to better customer experiences

I will remind @BrianHForr's point: target 30/70. 30% defensive, 70% offensive

Compliance is more than Data Privacy. #DataOps provides guidelines for regulations in general - and how they can automatically be understood for data requirements

Caitlin Halferty

an org-wide compliance scorecard can really help get all business areas using the same data source and racing to meet requirements

Dave Vellante30

Hello everyone - I'm your host today - looking forward to the chat - we'll be getting started right around the top of the hour...Where's everyone chatting in from today?

[+] Show Hidden Comments

Sunny Dublin, Ireland!

Toronto Canada

sunny Boston

Hello everyone. We are excited to have you all here!

Boston, MA!

Joining from North Carolina. Excited to learn more about DataOps!

Austin, TX!

Jay Limburn 🇪🇺

Southampton, UK

@ElaineHanley I need the most votes for Ireland :-)

Itumeleng Monale

A chilly evening in Johannesburg, South Africa!

A humid Toronto

Reading, UK

Hello from Palo Alto California sheltering in place thanks for organizing this chat looks like a great panel of experts and friends

Hello John - thanks for joining

Sunny San Diego CA!

welcome ti the chat @BrianHForr !

Bradley Shimmin

Longmeadow MA here folks. Greetings!

Paul Zikopoulos

@furrier Hey guys ... Paul Zikopoulos from TORONTO! (Though we have California weather right now)

welcome Bradley!

@TumiMonale Dumela, o tsogile jang! Do you speak Setswana? Many years ago I lived for 2 years in Botswana as a teacher

Streamlining data pipelines is the core of the practice, which we will discuss in detail.

Delay means delay in achieving business objectives

Patrick Gaffney

Hello from sunny Boston, MA! Here with the Forrester team @daveeitler and @BrianHForr!

welcome Patrick and the Forrester team!

Promoting success is so key to MOBILIZE the stale talent too.

Itumeleng Monale

@ChrisBergh That is my home language! :-) Re tsogile Rra. Malebo

Austin, Texas