There is an unprecedented wave of speculation and investment in Artificial intelligence (AI) sweeping the industry with ideas like machine learning (ML) appearing as a magic wand of business innovation. Beneath the seemingly boundless hype of AI, however, there is something quite simple at play -- the idea of finding ways to perform well-understood, common tasks more efficiently, according to leading data and analytics company GlobalData.
An analysis of GlobalData’s Disruptor Tech Database reveals that enterprise AI practitioners are primarily seeing very pointed benefits within problems that directly impact the bottom line.
For instance, one of the compelling areas where businesses have been using AI is generating leads. Harley-Davidson in New York chose AI to automatically craft the most appropriate digital marketing and advertising campaign on a customer-by-customer basis. The technology captures existing customer data from the company’s customer relationship management (CRM) system and analyzes user’s online and offline past purchasing behavior to scale up marketing campaigns across channels. Within three months of deployment, leads of the dealership grew by nearly 3,000% and more importantly the sale of motorcycles by 40%.
Another, pointed, pragmatic use cases for AI can be found in boosting operating efficiency. General Electric (GE), for instance, unlocked millions in cost savings using AI. Given its long operating record in multiple sectors and numerous enterprise resource planning (ERP) systems, GE has often faced challenges in arriving at a conclusion on its overall expenditure. AI enabled GE to combine all its 270 separate ERP systems into a single platform. The company realized more than $100 million in return on investment in different ways, including optimizing sourcing strategies, renegotiating contract terms, identifying cross-selling opportunities and reducing landing cost of products.
AI has been increasingly touted by enterprises as a key to customer personalization. Spotify stands a classic example in using AI to facilitate its customers with personalized music listening experience. Although players in the music streaming business such as Apple and Pandora offer curated playlists, they often fail to match individual listener’s music taste over time. Spotify’s “Discover Weekly” feature, curated algorithmically, became a sensation in personalizing music playlists more accurately at exceptionally large scale.
AI is of course the game changer of risk management to businesses in many ways. Online money transfers and payments company PayPal moved from using linear models to neural networks with deep learning (DL) for analyzing money transactions in real-time. The advanced platform helps to create scenarios related to positive and negative user behavior and contribute to improving the accurace of fraud detection over time. PayPal claims to have reduced the fraud rate to 0.32% of its total revenue, as compared to many peers at 1.32%.
The aforementioned are just a few in the expanding list of companies realizing tangible benefits with the use of AI.
“The operationalization of AI has allowed nearly every enterprise to grow smart, leveraging AI not globally but very specifically in solving well-understood problems, all without having to invest heavily in data sciences. However, while many are witnessing desirable results, potential AI practitioners should approach AI with caution, carefully weighing internal expertise against business needs,” concludes Brad Shimmin, Service Director, Global Technology and Services at GlobalData.
For more on how AI is being used in the AEC industry, and how the AEC industry can look to other sectors for how to best capitalize on the technology, check out BD+C's article, "Say 'Hello' to erudite machines," in the August issue.
Related Stories
BIM and Information Technology | Jul 25, 2016
Autodesk’s LIVE turns designs into video game-like experiences
Users can adjust navigation points, render styles, and even the time of day, with fluid and quick controls.
Building Tech | Jul 14, 2016
Delegates attending political conventions shouldn’t need to ask ‘Can you hear me now?’
Each venue is equipped with DAS technology that extends the building’s wireless coverage.
BIM and Information Technology | Jun 14, 2016
Autodesk and Trimble will share APIs to develop products that improve user workflow
Data and document management is likely to benefit the soonest.
BIM and Information Technology | Jun 13, 2016
The race to digitize the globe with 3D imagery
Tech firms are creating a highly-detailed virtual planet available instantly for those who would like to scrutinize it. SmithGroupJJR's Stephen Conschafter details the new technologies being used to map our world.
BIM and Information Technology | Jun 7, 2016
Conquer computational design: 5 tips for starting your journey
Data-driven design expert Nathan Miller offers helpful advice for getting your firm ready to use CD tools and concepts.
BIM and Information Technology | Jun 7, 2016
6 ways smart AEC firms are using computational design methods
Rapid prototyping, custom plug-ins, and data dashboards are among the common applications for computational design.
BIM and Information Technology | May 20, 2016
AIA and Autodesk introduce new feature to automate 2030 Commitment reporting data
The new automated connection will allow the more than 350 AIA 2030 committed firms to report their project and portfolio performance to the DDx directly from Autodesk Insight 360.
AEC Tech | May 10, 2016
Thornton Tomasetti launches new tech company
TTWiiN initially features six products and will add more via its own incubator.
Sponsored | BIM and Information Technology | May 10, 2016
Advanced laser scanning technology supports data collection and modeling efforts for Missouri’s Iatan 1 Power Plant
For the installation of a new heat exchanger, the power division of Black & Veatch contracted an engineering firm to laser scan the site, make a piping model in Autodesk® Revit®, and export it into AutoCAD® to deliver results.
AEC Tech | May 9, 2016
Is the nation’s grand tech boom really an innovation funk?
Despite popular belief, the country is not in a great age of technological and digital innovation, at least when compared to the last great innovation era (1870-1970).