Hype or Not? Proven Value from Artificial Intelligence
My new favorite vlog to follow is that of Simone Giertz. She is a Swedish inventor, maker and a robotics enthusiast. She is your real life, modern and feminine version of Dr. Emmet Brown (from the movie Back to the Future). Even better, she builds artificial intelligence/robots to help her with the mundane daily tasks like brushing teeth or making a bowl of cereal. I doubt that these robots are truly useful, but they are funny. Check them out.
In the world of Data Management, there are a lot of repeatable and mundane tasks that take time from your human workforce and expose your business to flaws. Developed and employed more seriously, Artificial Intelligence can make these tasks go away and in doing so transform your business.
Imagine the task of managing your trading partner community (For example – a retail supply-chain of order-to-cash or insurance claim process). For instance, if on-boarding a new trading partner network community used to take an average of 2 months with emails and faxes and 3 weeks once a B2B solution was deployed (such as Informatica Data Exchange), it can now be reduced to 2 days using machine learning algorithms. Informatica’s AI/Machine Learning capability, the CLAIRETM engine can assist the community manager/partner coordinator with the loading, parsing and mapping of non-standard files (such as XLS, CSV and TXT). Not only that, if the partner modifies the file, the process won’t fail. CLAIRE is smart enough to dynamically adapt to changes in file formats allowing the business to focus on doing business with trading partners versus managing IT processes.
We have seen similar gains in implementing data quality solutions. Informatica has been a recognized leader in data quality solutions to cleanse and standardize addresses, emails, names and other entities throughout enterprise systems for well over a decade (click here for Gartner report). Now, the latest version of Informatica Data Quality uses Name Entity Recognition (NER). In one customer’s Proof of Concept, we decided to test our hypothesis that using CLAIRE will bring greater value to our customers. We employed the standard deterministic methodology using reference tables, and then we did the same, only this time using CLAIRE. We ran a comparison: Development time for the deterministic rules was about 6 weeks with an end quality of about 85%. Using CLAIRE, we constructed a model in 4 days and obtained a result with 95% accuracy on the data set. The productivity gain was remarkable.
These are just two examples where we saw real cost saving and productivity gains that can be immediately measured. But, there is also the value that is gained over time. Applying Artificial Intelligence to data management changes the way you do business:
- Your data is more accurate – so you are less exposed to business process errors
- Your data moves inside and outside your organization faster – so you are less exposed to data latency
- Your data is whole – so you have a full view of your business and your market
For companies that are further along the journey towards data-driven digital transformation, researchers found that the impact is wide and real. For example:
- Employee engagement is higher
- 37% of leaders surveyed cited positive impact on employee morale
- 28% saw increase in customer satisfaction KPIs
- 41% reported increase in market share due to the digital transformation efforts[i]