International Institute for Analytics

Many organizations struggle with implementing agile or don’t see the expected results because they treat agile as just another process. If you are considering the use of agile development for analytics, it is critical to remember that agile is much more than a process – it is a philosophical and cultural approach to delivering customer value.

IIA’s 10th Analytics Symposium was held in Atlanta on October 10, 2018. This Symposium brought together analytics leaders from different industries, functions, and geographies to share insights and best practices. Experienced and innovative analytical leaders from large enterprises (Ford, Honeywell, Optum, GE, and Symantec/ID Analytics) described their organizations’ analytical journeys and key lessons learned—about creating business value, building an analytics team, creating a more analytically savvy company, and dealing with the hype around AI. Also sharing insights and trends were leading academics (IIA co-founder Tom Davenport and Jennifer Priestley) and IIA leaders Bill Franks and Kathy Koontz. Read the event summary for key themes and insights from the conference.

From Black Box to Glass Box and Other AI Trends

With automated AI-decisioning rapidly impacting society and our lives, it is becoming increasingly important to understand the underlying factors driving these decisions. Sameer’s presentation looks at some recent promising research into XAI (Explainable Artificial Intelligence) that ID Analytics is actively working with. He spotlights a couple of credible techniques that are enabling interpretation of complex machine learning models (‘black boxes’). The presentation touches upon the AI boom (and hype!) as well as ethical considerations when it comes to AI decisioning.

Competing on Analytics is a “Team Sport,” but Which Sport?

Alex Barclay, SVP at Optum (part of United Health Group, a Fortune 10 company), outlines the capabilities required to compete on analytics and describe how the company is boosting these capabilities among executives, managers and analytics professionals.

BI isn’t Analytics: Honeywell Aerospace Analytics Team Takes Flight

Beginning nearly two years ago Honeywell Aerospace had the common misconception that BI equaled the full potential of what analytics can accomplish. In order to find their current success, multiple fronts had to be established: educating the business side, finding executive sponsors, identifying projects worth chasing, and of course successful execution against those projects. In this talk, Dr. Seth reviews the challenges faced as he established an analytics practice and shares some of the specific case studies indicative of this evolution in analytics adoption for various business functions and competing on analytics in the aviation industry through new connected analytics offerings.

Model Performance Monitoring – Translating Models into Business Effectiveness

We’ve all been told that translating analytics work into tangible business value is one of the keys to empower successful analytics organizations. The question is how to actually operationalize that and scale across the organization. This talk will provide a case study on the business processes that Adam and his team have developed to define and track KPIs across multiple industries. It will give examples of how model results are then mapped to those KPIs and overview technological approaches for the continuous tracking of those metrics.

Inquiry:

One of the roles of our digital research team is to look at trends in the analytics space. What are the trends in data and analytics that we should be aware of?

Inquiry:

We’re in the digital entertainment industry, which requires high responsiveness to feedback so that we can continually improve the user experience. What investments can we make that will scale, particularly in terms of supporting real-time analytics?

This fall IIA is bringing together some of the top analytics professionals from around the country to speak at the 2018 Analytics Symposium – Atlanta.

Inquiry:

We have a good way to measure each customer conversation, but are interested to learn how best to enrich the feedback loop for the natural language processing (NLP) algorithms/machine learning models.