ChicagoBoy11

Many of you probably remember the company "Magic" from several years ago, which made waves by being a concierge service whose entire customer-facing experience was simply a phone number that you could simply text and make all sorts of requests from.

I work at a small consumer facing company but for our product most of our customers need to interface with lots of different folks internally. The "Magic" experience came to mind, and we were wondering if we could try and recreate it -- basically an efficient way to deal with so many disparate user requests but have it all, from the customer perspective, feel as if they only needed one simple channel to communicate with.

With that in mind, anyone know of any particular tech (be it existing CRMs, custom solutions, etc.) that the Magic team may have leveraged to accomplish what they were doing?

As a new father, I've found it extremely blissful to spend the evenings reading to my child. As he grows older, the one thing I wish existed was a way to much more readily indicate the actual words I was reading to him while still not requiring me to point to every word. Almost as if magically as I read, the word I was reading would have a soft glow around the book to just slightly draw his attention to this.

Thought maybe this would be a fun project to hack together. I've seen a few of the Arduino-controlled laser/water gun projects to not be so worried about that. What I am less certain about is what the best path forward would be for OCR of the footage of the device, coupled with plane detection of the page to estimate the pose of the book on my lap and properly do the trigonometry for the light/laser that gets shined onto the book. If anyone has seem any projects that leverage OCR/voice recognition and microcontroller robotics that I could learn something from as I try this out, I'd greatly appreciate it!