Home Bots & Business Blog Asher Lake: Simplifying the Application of AI

Blog Asher Lake: Simplifying the Application of AI

by Guest

Let me make one thing clear upfront: I am no psychologist. I am not a trendwatcher. But I am interested in how we, as humans, are reacting to hypes. When I was a little boy I experienced some hypes but due to the fact that my parents were immune to hypes somehow, I did not really care. Yes, I did collect stamps, I did like to play with marbles. But, Flippo’s, Pokemon, Tamagotchi did not really make my heart tick. Those hypes did not touch me. However, technology does. But not for the sake of it, I like to be pragmatic about it: let’s put it to work for us in a meaningful way.

Demystify AI

I am also not a data scientist. But I do love the potential of Artificial Intelligence. Over the past years I learned a lot of acronyms and terms that are connected to the topic: NLP, NLU, Computer Vision, CA, ML, Neural Networks, DL, Deep Fake videos, Vision AI and so on. And yes, I also learned why I should have paid more attention during my math classes. But is all of this a hype? It all comes down to the real value. Is it just big tech companies like Facebook, Netflix, Google, Amazon, AliBaba and so that can benefit from it? Or can every company reap the benefits?

AI seems to be the logical next step. And like moving RPA out of the hype zone we should do the same with AI. Demystify AI and make it applicable for everyday work at any organization.

Give Your RPA Solution Digital Brains

It would be great if we could all benefit. So let’s dive into a case that is worth your while if you have a lot of incoming traffic, like messages, e-mail, and even paper mail that you need to act upon. You could call it Intelligent Mail Automation or Intelligent Document Recognition. By combining AI models to understand unstructured text and categorize it for you, the next step RPA can help you out with is by uploading and logging the incoming messages and documents in your back-office systems. Maybe it could even manage the request from the customer end-to-end.

An example could be an airline: How did they process thousands of incoming voucher and refund requests at the start of the Covid pandemic? Intelligent Mail Automation came to the rescue with these 7 steps:

  • Step 1: The AI solution (a Machine Learning Model) was fed with all the emails coming in from passengers;
  • Step 2: The Model categorized the message based on how it was trained and was able to find the incoming mails that were about voucher requests or refund requests;
  • Step 3: The Model also looked up the passenger information, reservation number, or any other related information needed to feed the RPA part of the solution with;
  • Step 4: The RPA solution was triggered by the AI solution with the proper information;
  • Step 5: The RPA solution searches in all the systems of the airline to determine if the passenger was maybe already served via another channel, if yes the passenger received an automatically generated message in their mail with the results;
  • Step 6: If the RPA solution determined that no service was registered yet it collects all the relevant  data offering it to a Customer Service Agent enabling him or her to provide the requested services fast because all was prepared by the RPA solution;
  • Step 7: The Customer Service Agent decides what’s the best service to offer, connects to the customer with an offer. Upon approval, the Customer Service Agent forwards that decision to the last part of the RPA solution which is then able to register, administer and inform the passenger about their request offering them a voucher or refund as requested.

Do you recognize the patterns? This also works when you are an insurance company and want to check and process expense declarations from your customers. Or if you are a government agency and want to process incoming messages from your citizens.

You get it, right?

So start your AI game today. Start simple, by adding brains to your automation.

Asher Lake is the founder of Unlocking Digital

Originally published by Ciphix

Misschien vind je deze berichten ook interessant