Intermediate intelligent automation

After I shared with you in the last post how I started to automate small ideas I had related to how I personally my device and its vast internet applications, now it is time to move on and explain the next few ideas I had, which by now gathered a bit more complexity and business intelligence in them.

I mentioned already the ingelligent automation job aggregator. I did the insane task of building over 360 bots, one for each company that I found online and had a Careers page. The bot bot would go to the careers page of that company and type in the search bar the word “RPA” and then extract the numebr of vacancies the query results in. I extracted that number and then I have put together a bot that was sharing the weekly graphic and percentage change in Facebook and LinkedIn groups.

Which leads me to the next intermediate script with a lot of potential gain for your social life. I have build bots that do marketing a message. By marketing I mean sharing the message in one on one private message, instead of posting it in a public format on the same social media platform. There are huge advantages of one on one messages compared to the standard public post. People react differently to one on one messages than to a public message, even though it is from the same person. You will get much fewer reactions, however, you will also get some, though less, bad reactions. The postitive reactions are much more important here than at the public post because these are the honest reactions compared to the likes or comments of the public post. This is very valuable tool.

The next idea I had was to post messages on blogs in the sections comments. This did not have any practical effect than me being banned from those blogs, but at least this is in theory a cheap way of marketing as well. Some of the blog owners accepted the spam and I got some extra genuine organic traffic like that.

Next was an idea with a lot of profit potential. Inspired by boyish activities of sports betting, I have found a concept called value betting. Value betting means finding odds that are better than the real odds of a sports event, and bet as many bets possible on these value bets. Statistically speaking, you will tend, with a high number of bets, to profit exactly the percentage above the real odds of the event, in this case around 3%. There are really high fee companies which do this for you, but as I said, the fee is killing. So, I have encountered a site called oddsportal.com, which aggregated the odds for the same sports events from tens of different bookmakers. That was enough for me to extract those odds, and just subtract the odds of bookies I had an account on from the “real” odds of the event, which by tradition is the odd of serious bookmakers like Pinnacle or Betfair. Very easy calculation and voila, there were my value bets. For free. In theory this would have led to a profit, however after 800 bets I realised I am going negative. This can happen and in the value betting community it is called variance. So I quit that.

Then, also related to sports betting, I have built statistical models based on sports match statistics to predict the winner of the game, or other type of bets like number of points, goals, etc. In order to do that, I built a bot which went to a site called flashscore.com, and extracted all the stats from all the games in a season. Based on those models, I was betting on both team to score or under or over. Until I realised that I am not winning anymore. However, there were discoveries I made while losing money, and these might be foundations for winning strategies in the future. For example, I have found a winning streak in April which was related to the fact that teams score more in the last games of the season. More than the bookmakers have expected.

Another automation idea that I am really passionate about is the robot DJ. The robot DJ is an automation in Traktor PRO that mixes some tracks you feed the bot. The mix is done automatically, with the bot pressing sync and adding the next track to the decks. I find this really interesting as you can go to a party and instead of focusing on the mix, you can actually focus on the socialization part, pretending just that you mix the tracks. What can be more enjoyable than being center of attention, being paid for it, and you enjoying all the attention instead of working hard? Imagine that those DJ are doing repetitive stuff there. If they would automate most of their job, their DJ focus would go on the track selection more than on the actual mixing, which now takes too much of their time.

Another great idea I had was the image downloader from Getty Images. I have a passion for UNESCO World Heritage sites. And while working for one of the biggest corporations, I had the privilege of getting a corporate Getty Images account. This perk, together with my passion for automation and great list of places around the world has resulted in a bot that searched for the city where or near where a UNESCO site is located in Getty Images, and downloading the most popular picture of that place. This has resulted in a stunning collection of photographies that I will explain to you in another post what has been the basis of. Art!

Then, I have the ideas of an automation which searches for my next book to read. Since 15, I have started reading massively. I have at least 730 books read in the last 10 years, and still counting. I have a list of over 8000 titles saves in my Google Chrome bookmarks, 99% of them non-fiction, just like the 700+ which I have already read. And this bot that I am going to present now reads the bookmarks and then goes to the National Library of Luxembourg catalog and searches if the book in my bookmarks (with Amazon.com URLs) are phisically available in the NBL catalog. Bots for culture as well!

I hope I could give you some ideas on what is the true power of intelligent automation for non-technical people who want to skyrocket their intelligent use of their device, instead of watching 9gag or playing stupid game or reading governmental propaganda online!