PODCAST Artificial Connect uses AI to generate automated local event announcements
This story was migrated from our old blog, originally published on November 12th, 2020.
In Hannover [a city of 790.000] there are 40.000 to 60.000 events in a normal year. The problem of the publishers was to first generate all the event information and second to write articles about them.
Felix Rolf, CEO Artificial Connect
Media Partnership
This interview is in media partnership with the CONTENTshif accelerator program (https://www.contentshift.de/en/accelerator/programme/for-startups/), which we follow since its inception. CONTENTshift is the accelerator program of the German Book Publishers and Printers Association. Below you will find more interviews from past batches. We used to record the interviews directly at Frankfurt Book Fair, but since it is canceled this year due to Corona, we resorted to remote only interviews. At the time of the recording, we did not know who won the final award. The winner of this year’s batch is SciFlow (https://www.contentshift.de/en/contentshift/news/congratulations-to-sciflow/). We will publish the exclusive interview with them as the last of our series this year.
The Founder
In this interview, we are talking to Felix Rolf (https://www.linkedin.com/in/felix-rolf-526a4894/) the founder and CEO of Artificial Connect (https://www.artificial-connect.com/), an AI startup located in the beautiful city of Hannover.
Felix originally started with an advertising conglomerate with roots in newspaper publishing, where he got to work in several departments. He enjoyed his time in digitalization projects, especially in data science. During his time there he worked at the intersection of data science and sales.
I participated in a boot camp for three months to learn about the statistical tools needed for data science.
Felix Rolf, CEO Artificial Connect
Affiliate Links
Why not learn from entrepreneurs and executives who are open to sharing their favorite books. We skimmed through some recommendations and that is what we found.
- Recommended by: Bill Gates
“Enlightenment Now by Steven Pinker” https://amzn.to/2XAENlB - Recommended by: Jeff Bezos
“Built to Last: Successful Habits of Visionary Companies by Jim Collins and Jerry Porras” https://amzn.to/2RGKGKe - Recommended by: Elon Musk
“Zero to One: Notes on Startups, or How to Build the Future by Peter Thiel” https://amzn.to/2REbHhw - Recommended by: Tony Hsieh
“Tribal Leadership: Leveraging Natural Groups to Build a Thriving Organization by Dave Logan” https://amzn.to/34KgraQ - Recommended by: Warren Buffet
“Business Adventures: Twelve Classic Tales From the World of Wall Street by John Brooks” https://amzn.to/2K92z06 - Recommended by: Tim Cook
“Competing Against Time by George Stalk and Thomas Hout” https://amzn.to/3cjMMYn
The Startup
The story of Artificial Connect (https://www.artificial-connect.com/) starts with Felix’s fascination for data science. He taught himself Python and participated in a 3-month course to learn about the statistical tools needed. Of course, as in many good stories, there is wine at the beginning. Felix wanted to help an online wine shop make wine recommendations, so over a bottle of chardonnay, they decided to go ahead with the project. Together with the encouragement of his mentor from the statistics course, he decided to go ahead.
As a start, they worked as a consultancy mainly for media companies, especially local newspapers. In the local newspapers’ home market, the local staff had problems with the coverage of local events. After hearing about the problem from two different clients, they decided to apply their accumulated knowledge to local events.
With a bit of adjustment, Artificial Connect can apply their tool to other areas, where they are already in discussion with several other industries to widen their scope.
We divide all the events in a long tail and a short tail. The short tail there are the large concert. In the long tail there are many local players, which no one covers.
Felix Rolf, CEO Artificial Connect
Venture Capital Funding
Artificial Connect is open to external investors and is currently bootstrapped, without an external investor.
The Video Interview is set to go live on November 12th, 2020
The Audio Podcast
Further Readings / Additional Ressources
Hannover https://en.wikipedia.org/wiki/Hanover
Hannover Dialect (high German) https://en.wikipedia.org/wiki/High_German_languages
Integrated studies (in German known as Duales Studium) https://www.alumniportal-deutschland.org/en/study-continuing-education/study-training/dual-study-programmes-germany/#:~:text=A%20study%20programme%20that%20offers,commercial%20qualification%20Industriekaufmann%2FIndustriekauffrau)
Feedback
Reach out to us, here is our audience survey, to give us feedback, suggest topics, interview partners or just to say “Hallo!” https://forms.gle/mLV6mVKwGwKuut8BA
The Interviewer
This interview was conducted by Jörn “Joe” Menninger, startup scout, founder and host of Startuprad.io. Reach out to him:
LinkedIn
Twitter
Email
Automated Transcript below
[0:00] Music.
[0:08] This is Joe from Startuprad.io your startup podcast and you to block from Germany
today bringing you another interview in our media partnership with the CONTENTshift accelerator program
which is the accelerator program of the German Association of book printers and book publishers nonetheless we don’t have an actual book printer or publisher here in the interview of course we do have a start-up here but first I would like
to welcome the CEO and founder Felix.
The Founder
[0:53] Thanks hi everyone it is a great pleasure to have you here and you guys are called artificial connect Wheels pretty soon get to that.
Let’s first go a little bit through your life and what you’ve done which led you to this stage
actually, the funny thing is you are from Hanover yes there’s a Hanover in Germany for all the Americans out there and yes it’s a very nice place I’ve been living there for a project for quite some time and I’m pretty familiar with the area around much see
for everybody who knows a little bit German the official German language,
without any dialect was one at one point decided to be the Hanover dialect so you speak proper High German right.
[1:41] Yeah I hope so but I’m not really sure if I can transfer that to English so the English listeners would not hear it that yeah everybody says so that we might speak very clear German.
huh good place to start if you want to study German that said I’ve been stalking you are on LinkedIn like I do put all my guests and if you guys
if you,
viewing this on YouTube or if you are listening to this go down here in listening to this in our audio podcast go down here in the show notes you will find a link to our blog post which will contain
your LinkedIn profile that said if you if you’re new to our Channel hit the like And subscribe button and be sure to ring the bell that you don’t miss another episode that’s it,
you actually started out an advertisement how did you end up with an AI company.
[2:37] Well I started in 2012 working for the mud sack media.
That’s a company quite a big company based in Hanover with subsidiaries in a lot of cities-it’s a media company formerly a newspaper company
but after that, they went into a lot of
other companies and industries and I did a corporate study program they are in Germany it is quite well-known I think,
for the listeners outside of Germany are outside of Europe that’s a program in which you can study and work,
at the same time so a company is funding your studies.
[3:34] And I did that in Hanover with Matt sack
and went into different areas of the company Finance controlling sales marketing online marketing,
strategy and I really liked the sales department there
because of several reasons because of the strong focus on National customers on digital sales online marketing
and there were several digitization projects hat.
[4:10] And that I was able to participate in so I worked after this three-year program.
[4:19] That I worked on from 2012 to 2015 I work for about one year and the national key account management
with big customers from
from different industries of Germany and from that on I went into some other departments.
Of Matt sack that included especially the strategy department where I got into contact with a lot of data projects since I studied economics
my bachelor’s degree that was to some degree quite new especially on the operational side.
Um I was able to learn a lot during that this projects and
then another project occurred at
the interception of sales and digitization especially to if it comes to intelligence systems that might predict,
which customer you should call at which time with which
so that is how I got into more contact with data.
And sales and I did that until 2018,
so that that has been my maps that years.
[5:45] We may add for everybody who’s not familiar with the integrated studies as it’s sometimes called basically it’s you go to college.
June semester and during your time off when you don’t have classes you’re actually working and during the whole time you get the same salary it’s a very interesting concept here in Germany.
Getting Interested In Data Science
[6:08] So you have been working in a company where you also studied and then you got in touch with more and more data science in relation to sales what did it do with you what did you learn from there.
[6:23] I learned a lot about the data structure about how you can handle big volumes of data.
And I got an understanding not a deep understanding because I wasn’t able to program to develop on my own,
but I got an understanding on how data infrastructure works and on how software based products,
are functioning at its core.
[6:57] And since I wasn’t able to understand it I didn’t know about any any program.
Yeah any coding any programming languages I decided to learn that and then I started started learning python especially for data science.
That sounds like a really really busy time well
I started learning Python and I started to to understand the syntax and the idea of Pi c as an object oriented programming language,
but that was only the one side the other side was,
all the statistical background that I needed to cut into moderator science and then I decided to take part at a boot camp.
Based in Hamburg it’s called Noah Fisher I think boot camps like that are accessible all over the world so far.
[8:07] I’ll buy now and then this boot camp I spent three months.
On developing basic data science skills we had a very very high skilled coach.
It’s name is Matias and Matias worked as a chief data scientist before.
[8:29] He worked at Zalando that’s a landowner Tech Lab he is a practitioner through and through he once studied mathematics and then did his Ph.D,
in Applied Mathematics and then.
Um work for several years at a young company which they later sold to Solano and he always worked in the data Department.
And the data science handle data science tasks,
and for me that was very interesting because I had a very very good coach who knew his stuff and who brought us all this data science knowledge,
on the table through new cases every one or two days from the business side and in this,
three months period i extended it a little bit too far off I funds but in this,
main three months period I kind of got the entry to data science and learn the basic knowledge.
The Startup — From Wine To Local Events
[9:41] I see and what was the moment you actually decided to make.
Your own company of the other one was the moment when years later hmm that could be something.
Out of which I could turn into a real business what was that moment like.
[10:02] Well I think there have been two moments the first is I.
Besides coding I have a passion for wine and I started last year to become some of you.
Um and I studied for that and I learned a lot about wine about wineyards about the different styles in harvesting the wine at the grapes and wine stocks,
and during that time I got into contact with an e-commerce company that is selling wine online.
And they told me about the problem they had.
They discussed they discussed how to implement the recommendation engine.
So that you as a maybe as a wine enthusiast or a potential customer.
Getting some recommendations on which ones you might like and you might dislike.
And we we thought about that together I think we had to shut down a doing that,
and while we thought about that I.
[11:19] We came to the idea that we should just start to to develop such an recommendation engine.
And there are several types which we then developed and that was a point from which on I thought well,
I really like that especially if it comes to to to tasks where you can automate a processes that generate higher value for companies,
like left natural language processing test for example that was the first point and the second the second,
point that occurred was I spoke a lot with my former head coach Matias and with.
Another person who attended this this boot camp his name is Alex we spoke a lot about
business ideas about machine learning ideas and not only ideas but especially neat which problems are not solved yet and are there some problems that you might,
solve with machine learning and then we.
[12:31] We thought a lot about natural language processing about media about especially about the generating of content.
And how did you end up from the very fun area of recommending wine to doing like something like automated.
Very localized event announcements.
[13:05] Um so since Mark Lagos the defaults founder
co-founder of the artificial connect and I have a background in the media industry we had a lot of contact contacts too
two media companies and the first project we did which were,
mainly Consulting projects we did with those media companies so we were in the Consulting business and we
even a developed some some services for them some smaller prototypes or some some products that they are using now
and within this meetings we spoke with a lot of.
[13:54] Of newspapers especially with the editorial staff and they told us that they had several problems in there.
In the in the local market,
especially if it comes to local journalism and to the event market and the first time we heard that we thought well that,
that’s interesting and then we I think we went on with our other projects and the second time we heard another customer telling us about the same problem,
we got more and more interested in that and then we asked a lot of media companies in Germany,
what they are thinking about this specific problem and if we would solve that problem would that be a high would that provide a high value,
if we would develop an automated solution for this specific task.
So most of of the directors and so on we spoke to.
[15:02] They said that the really would be a great value for them and then we looked at on the business side,
on the financials and after a while we decided to go into that market the market for structured text generation,
where we are building a machine learning product.
That is able to automatically generate content based on just a few input words or input sentences that you are giving,
the product.
Basically you have the rough idea of a recommendation engine which turned into.
From a wine very fun,
every recommendation into and also fund like very localized event recommendation is that about true.
[16:02] I think that’s one side of it but it’s.
It’s only the tip of the iceberg the the main problem we would like to solve is if you’re looking at the local market in every city in every country,
in a time in which you are not speaking about Corona,
in a time where everybody can enjoy themselves theirselves and go out and have fun we are seeing a lot of events especially in the local market
events that you have to pay for to go there or free events.
That do not have a price to to enter and this event Market there are several a lot of players included actually.
[16:49] We are separating that that event marketed to a short tail and 2 into a long tail.
The hotel we would assume that there are events like very big concerts or smaller concerts but events that are quite well-known events that you have to pay for in general,
and events that are covered by a lot of ticketing companies.
And then we have the long tail and in the long tail there are also events that you have to pay for to go there but there are also events that you do not have to pay for that are part of the of the cultural life.
And this long tail event Market is very fragmented it consists out of a lot of players with in this market.
And there’s no player that covers all the events.
[17:42] And one problem that all the Publishers that we worked with together hat was first to generate all the.
Event information about all the events that are going on in their local neighborhood.
And second to White articles about that specific event,
about all this events an example from from my home town from Hanover is there are between 40 to 60 thousand events each year.
And the only source where you can really find all of those events in an inn.
Some kind of open open source is newspaper.
But the newspaper is not writing an article for every event maybe only for 5 to 10% and that’s very cost intense and what we are developing as a service.
[18:40] Which generates automatically event announcements so a service that first.
That first the others all the event information and then take for example the title,
the plays the prize the host and so on uses this information categorizes it classic classifies it and and then generate articles out of it.
That you can publish in a print medium or online to generate more visits and to provide a better customer experience for example for your for your fans or for your listeners.
Or your readers,
and the third part then is to per to personalize the way in which we are presenting this event information or especially in which our customers,
for example newspapers are Publishers in the book industry are.
Distributing the event information through different channels.
[19:52] I see I see um before we get into what contentshift did for you I have one more question what type of areas are you currently thinking about like to also get in because I assume you wrote,
a solution for one problem but with a tiny bit of adjustment it may be able to be like apply to other problems as well.
[20:16] That’s right there are several markets that are quite interesting for example in general e-commerce and some niches of that.
And the whole pharmaceutical Market is quite interesting as well the Health Market in general.
And then there are several other Industries smaller Industries compared to e-commerce or the Health Market.
That might be of interest as well.
And areas in which we are currently speaking to potential Partners to enter those markets.
What would you do there.
[20:59] We would provide for example a service that generates texts and large volumes of text for example if you are an e-commerce company,
and Q have a you have a lot of products in your stock,
and you have product descriptions that are duplicate content for example
we could create unique content for example a company that sells a washing machines laundry machines are that sells wine online.
That’s that is selling other things.
We could provide them with with product descriptions that make them or that that I helping them,
to get a better search index and so on.
[21:49] Hmm I see see see one more question how are you guys currently Finance are you currently looking for external investors.
[22:00] Um we are wide open we’re quite open to have conversations about external Investments so far we chose to bootstrap the whole company.
And until this point we do not have an external investor.
Contentshift Experience
[22:20] I see you let’s get a little bit into our partner because first how did you learn about contentshift and secondly,
what did it do for you what was the benefit of participating in it.
[22:36] Yeah so a market that we are very interested in is the reading Market readings about books.
And I think readings are getting more and more popular.
And when we thought about the way that our products might fit into the book industry especially because we are interested in readings we,
we found out that.
The program contentshift helps young companies to develop their products and to data to give them a space in which they can have a very very.
Very intense discussion with persons from the book industry in a very high wall managers Publishers and so on.
To speak about the pros and cons the benefits of the product and where we should.
[23:43] In which way we should we should go and develop our product so we did some research and did not take us long to get in touch with contentshift
and then yeah we applied and took us a few weeks until we got,
message that we are in an offer which we were very happy and then three months program started,
it’s nearly over now so I think about three months ago it started and,
at first, we had a lot of possibilities to discuss our product within a mentoring program.
So each company had an at own mentor.
And this Mentor provided us with book industry insides we spoke about the product we both spoke about possible adjustments that we had to take,
to data to develop the product in a more adjustable way for the book industry.
[24:57] And then we had several meetings and workshops in which we presented
at this point current state of the product development and the next milestones we would like to reach.
By October and by the end of the year and we got a lot of.
[25:21] Very very good input there from from the people.
From the managers gathered inside contentshift.
So it was it was more than that work the feedback.
That you actually could approach the people in in the book and Publishing industry that there was the main benefit for you guys right.
[25:47] Data cement benefit for us and I think so far the program is not yet finished but so far it for us was a big success because we learned a lot about
how to really adjust the product and to reach a good Market fit.
[26:05] I see and we recording this actually knowing you are a finalist but we not sure if you guys have one.
But it will be published afterwards so if you go down here in the show notes you’ll most likely see our link to the winner only thing left for me to say is best of luck hope to talk to you soon again.
[26:30] Thanks thanks.
[26:33] Music.
[26:39] Interviews at www.starballroomdancestudio.com sharing is caring.
[26:47] Music.
