Applying collective intelligence to ML enabled data [LISTEN]

Copenhagen will bear witness to the return of IA Nordics for it's fourth installment March 18-19. While the lessons-learned have become more and more exceptional, the region has always showcased cutting-edge work. Here are some key insights from the past few years at IA Nordics.

Register now to join the show.

Buy, build or grow your AI talent

Ep. 119: Henrik Hanasand, Norway

"After relying too heavily on consultants, we decided to search internally for talent. It worked. Our new approach to a holistic IT solution has given us the flexibility to look at AI more broadly and futureproof their organization."

As noted in our respective ML Report and Preditive Analytics Report- you can buy, build or grow your talent. There are advantages and disadvantages to each approach. 

Digital transformation leads to geographic leadership

Ep. 118: Brian Mikkelsen, Denmark

"The launch of the Digital Growth Strategy for Denmark has kept Denmark’s economy competitive. Training the workforce of tomorrow through a focus on STEM education affords Denmark the opportunity to grow its economy locally. The government has established a digital hub, where small companies can have knowledge about AI, big data, and other new tech for free."

Denmark is on a digital transformation journey of it's own. Federal governments in general are further diving into next generation technologies as a way to attract job creation to their respective geographies. 

Profitable automation proliferation

Ep. 117: Alexander Hubel, Ericsson

"We've been working a lot with change management and communications to create automation communities throughout business units and group functions. From the automation communities we created something we call the automation tribe which is comprised of key decision makers."

Alexander goes on to say that these terms don't equate to a CoE as much as they do a cultural steering committee. His team has engagement leadswho are working more from a business transformation perspective with IT and the communities and tribes to ensure automation profitably proliferates the enterprise.

Build strong governance and engage across the enterprise

Episode 23: Susanne Skaarup, Danske Bank

“Start with the small processes, get some learning. Engage the organization and create awareness about robotics. Ensure that you have good monitoring of your robots. Build a strong governance around it and have your methodologies and standards in place from a central point of view to ensure the quality of the work.”

The above quote perfectly encapsulates what needs to be done. Get an understanding of what should happen specifically within your organization. Then get success under your belt. Then communicate, communicate and communicate. Susanne visited 80 departments to ensure that the entire enterprise understood the potential of intelligent automation and how it could benefit them. So Susanne emphatically can check the what’s in it for me box. On governance, whether your approach is centralized or decentralized, you still need to build a strong governance to ensure that best practice remains best practice.

Are they bought-in?

Episode 22: Alexander Hubel, Ericsson AB

“Secure executive sponsorship because with that funding comes access to executive leadership and it all becomes so much easier.”

Are you still trying to go at intelligent automation as a purely disruptive force within your enterprise? Notes from the field suggest that approach is fraught with danger. We’ve passed the inflexion point from ‘early days’ to ‘mass adoption.’ And so if you’ve not already done a PoC, then got a few bots working and ensuring you’ve secured executive sponsorship, you’re very much behind. And if you’re currently in purgatory- you’ve done all of the above but you still don’t have executive sponsorship- you’re falling behind. The good news is that all data points suggest that you don’t only save cost for the enterprise, you increase pace of work and you have the opportunity to innovate and ensure digital transformation is occurring through intelligent automation. But if you’re executive leadership is still kicking and screaming that they can’t see the light, check your message…and then if necessary, see what else might be out there for you. 

Causation not correlation

Ep.20: Michael Natusch, Prudential

"Where the cost of getting it wrong is high and the volumes are lower and maybe  where regulation is important- and we need to convey the reasoning behind what we do to a regulator or to somebody else- there we need to have some level of understanding of causation."

Michael's greater point was about AI recognizing patterns. And in some cases, messy pattern recognition through correlation works. But for enterprise, the stakes are much higher.

Applying collective intelligence (human + machine) to ML enabled data

Ep.16: Mattias Fras, Nordea

"We have robots that are screening customer data and then going out on the internet and finding additional information about these customers. What kind of companies are they involved with? What kind of social activities are they involved with? What kind of political activities are they involved with? So gathering data from external resources."

Are you harnessing the unstructured data available to your enterprise? Are you applying collective (human + machine) intelligence to ML enabled data?

Misunderstanding your data and/or processes

Episode 12: Mads Andersen, City of Copenhagen

“Our first foray into automation turned out being way more complex than we thought. It turned out that a small process involved 5 or 6 different legacy systems which the human connected.”

Sound familiar? Depending where you are on your journey, you have, are or will experience this exact scenario. There are a few reasons for failed automation implementations. But the two most prevalent are misunderstanding your data and/or misunderstanding your processes.

Business informs technology

Episode 10: Jorgen Lislerud, Circle K

“It’s important to have subject matter experts working with programming a robot – people who really understand what they’re doing. But we also need someone there to challenge the SMEs. That way when it’s put into the robot, we have the best possible input and optimize the process.”

Technology only works as well as the business not the other way around.