Intelligent Document Processing Podcast - Episode #3: Fabian Seimer, Partner and Head Information Management at Inacta

For the third episode, the CEO of Parashift, Alain Veuve, has welcomed one of his partners at Parashift’s office, Fabian Seimer, Partner and Head Information Management at Inacta.

They exchanged views on the current vision of their company and the future of IDP and digital transformation.

What about Inacta?

Inacta is an IT company founded in 2009 that accompanies companies in various sectors (insurance, banking, real estate, and health) towards a digital transition based on innovative and efficient technologies. Inacta’s experts accompany clients from project conception to completion, taking care to select the most efficient technologies and the most adapted to their needs. Indeed, the company has to offer its customers all the latest and most advanced solutions that can be found on the market.

With more than 80 experts in project management and software development, the company became a founding member of the Crypto Valley Association and the Swiss Blockchain Federation.

As an initiator of the cryptovalley.directory, the Blockchain Summit and the Blockchain Competition, it promotes technology start-ups and their networking with established companies. In this way, Inacta contributes to the continued economic prominence of Switzerland and Liechtenstein.

Inacta website: https://inacta.ch/

What about Parashift?

Parashift is a machine learning company that has taken on the topic of document capture. Its product, the Parashift Platform, is a cloud-based document extraction solution that uses IDP technology to read relevant data from any document. By using its proprietary Document Swarm Learning technology, companies can gradually automate all their document processes.

In the process, employees team up with the Parashift Platform to integrate document data into processes and systems super quickly and cost-effectively. The platform learns from all interactions and thus comes closer to complete autonomy step by step. So everything at Parashift is focused on making this technological milestone a reality. The team is not pursuing other topics. Alain Veuve says, “We’re not an end-to-end solution, so we’re working with a lot of companies that can do vertical integrations.”

Transcript

[00:00 - 00:18] Intro.

[00:18 - 00:34] Alain Veuve, CEO of Parashift: Welcome to the Intelligent Document Processing Podcast. I have with me today a person who has quite a lot of experience in the field. Fabian Seimer, welcome. Maybe say two or three things about yourself.

[00:34 - 01:00] Fabian Seimer, Partner, Head Information Management of Inacta: Thank you very much Alain for the invitation, I'm happy to be here with you. Yes, my name is Fabian Seimer, I'm a partner at Inacta, I'm responsible for Input Management and I've seen a lot in this function, I've been working in this area for about 10 years now. I've always been involved in larger projects for various clients, and yes, that's how we came into contact with each other.

[01:00 - 01:07] Alain Veuve: I like to remember when we...I think the first time we saw each other was at a summer party at your place....

[01:07 - 01:08] Fabian Seimer: Yeah, exactly, exactly.

[01:08 - 01:36] Alain Veuve: ...in Zug. And you came out of there and Thilo, the Head of Product at Parashift, who has known you for a while, said to me "This guy here, Fabian, he knows everything, he's seen it all, 'the bad, the ugly, the good'." And over time I realized that was true. You have an incredible amount of experience. How do you see the development of the last 10 years in terms of document processing? How have companies been mastering?

[01:36 - 02:45] Fabian Seimer: Yes, I think you have to look at it on the different industries. So, for example, if you look at the insurance industry or health insurance companies in particular, they started very early to digitize, but also to automate. And it's also clear that there are huge volumes of documents that have to be processed efficiently. This means that appropriate solutions are in place everywhere, highly optimized to achieve a high level of dark processing and so on. Unfortunately, other industries have not yet reached this stage. I think that in the meantime, a change has taken place in the sense that people are no longer looking at the large volumes, a central incoming mail is still an issue, but less and less so. It's now really about integration into the business processes. I think that is the big change that has now taken place. That is, it's no longer necessarily large volume that plays a role, but it plays a role in how I can actually use IDP in my business processes, so that there is added value.

[02:45 - 03:03] Alain Veuve: What we actually see here is that this whole RPA movement or almost RPA hype has actually brought quite a lot towards detail processes, so IDP into the detail processes. How do you see that? Do you experience that in your daily...?

[03:03 - 03:37] Fabian Seimer: Yes RPA is definitely a topic and in that context also more and more the corresponding IDP stories. But I would generalize it a little bit more. I experience more and more, it's about customer portals, it's about apps, it's about workflows, and in all these processes IDP plays a role. That means I always need to be able to quickly classify a document somewhere, extract data so that I can continue to work my process with that data. And with RPA, of course, it's exactly the same thing.

[03:37 - 04:10] Alain Veuve: If you go on there now, where is the pressure coming from that more and more is digitized. I may remember, I was still on the road as a consultant 10 years ago and we had to convince people, yes the digitalization. I talked quite often with people from executive management who told me, "Yes, digitization is nice, but it's not that necessary for us yet." I think that has changed. Where did this pressure come from? Or what triggered it, that we are already actually also operationally in 'full swing', so to speak?

[04:10 - 04:57] Fabian Seimer: Yes, of course that comes from the fact that the customers have also changed, the customer needs are quite different again today. There is insane competition, for example, in the health insurance environment, but also in the banking environment. I have a lot of competition there today, I have to be able to react quickly, I have to be able to respond quickly to customer inquiries, otherwise the customer is gone. I believe that a great deal has already changed, and this is leading to digitization projects and automation projects being launched everywhere, perhaps always on a small scale, related to a specific use case, but then going very very far. In other words, with real consistency from A to Z. And I think that is the reason why this is now playing an increasingly important role.

[04:57 - 05:05] Alain Veuve: So what you're saying is that it's actually more in the companies that you go forward in a use case specific way and you have little "plants"...

[05:05 - 05:54] Fabian Seimer: Exactly. And that was not possible in the past or, in the past I had these big input management solutions, you had to put a lot of effort into it to even install it, to configure it, so that I can even get started. In other words, such small use cases were not profitable, because if I only have 100 documents per year, it's simply not worth it to manage such huge projects. And that has definitely changed, thanks to solutions like Parashift, I can now also try things out in such small use cases, in such small POCs, perhaps at various points in the company. And we're definitely seeing that everywhere in the company, you called it "little plants," that those are emerging where you want to work with IDP.

[05:54 - 06:36] Alain Veuve: I also noticed, we were allowed to do some things together, and get to know each other a little bit better. You also have such a...how should I say this, without this coming across wrong now...a certain fascination with the things that can go wrong...So is meant in the best sense. I have often somehow in a project or with the customer a decision falls, which we can understand perhaps at first glance times not so well, you are someone who actually reacts to that with a certain humor and goes in there. Is that how you get when you're faced with external decisions for like 10 years through all this consulting work or are you just like that? 

[06:36 - 07:15] Fabian Seimer: I have both of course. I think certain things you can't change. And I think then let's see where we can do something together or where there are certain things that we can move, that we can push forward. Especially with big customers and in big projects, there are always situations where things happen that you can't influence. This has nothing at all to do with the content of the work in normal cases or with the solutions, there is simply also politics involved and so on. Yes, that's how I deal with it.

[07:15 - 08:03] Alain Veuve: I think that's good! Where do you think mistakes are often made, are there kind of patterns that repeat themselves, now especially in big companies, that projects fall behind, so a little bit...so earlier I would say, earlier when I was in eCommerce, there were often that projects like actually went wrong, so not the technical projects, but the business initiatives, you had to say at some point "Hey, that didn't pay off at all and we're not really doing that anymore." Today, I have the feeling that it still goes wrong, but not so fundamentally, it runs so partially off track and then you can save it again. What do I have to do as an entrepreneur, as a decision-maker in a large company, so that this doesn't happen to me? Do you have any lessons learned?

[08:03 - 09:18] Fabian Seimer: Yes, definitely what I experience in all digitization projects. The issue is mostly that these processes are simply overlapping. That is, I have documents and they may not be focused on a specific area, but they go across my entire organization. And that's something that I think for a lot of companies, for a lot of large companies, is something that they don't know. Projects are done in the specific units, they work very very well. When it becomes overarching, then the processes, the responsibilities for it, are missing in order to be able to map such cross-cutting issues. On the other hand, this means that what you see in the textbook is also true from my point of view. So digitization projects always have to be approved from the very top, but they also need the appropriate care, the appropriate attention on the subject, otherwise it doesn't work. Too many kingdoms ruin the project. That's really one of my most important learnings from past history.

[09:18 - 09:21] Alain Veuve: I would add at this point, budget would also be important.

[09:21 - 09:26] Fabian Seimer: Absolutely, absolutely. That helps.

[09:26 - 10:08] Alain Veuve: Yes, that helps. Okay, I think I can relate to that. That's so this cross-departmental digitization that often big companies have the cramp on. Now that would have been the classic role of the Chief Digital Officer. Once upon a time, a star rose up, and everyone had chief digital officers. I was always very critical of this, because I thought that digitization projects are actually change projects, and that this is something that is actually top-of-mind for top management. In Germany, in particular, I've seen one or two companies where they've said, "Yes, we'll hire someone to handle digitization, and they'll do it."

[10:08 - 10:41] Fabian Seimer: Yes, I think that is also the easiest way from the company's point of view. I have someone who takes care of this topic in a dedicated way and that person then makes sure, that person then makes sure that digitalization is pushed forward in the company. But yes, as you say, digitization is a change topic and that affects everyone and that affects the entire management team. That is, I also have a bit of a hard time with this role of the CDO or whatever they are called.

[10:41 - 10:46] Alain Veuve: You don't see them that often anymore, I have the feeling. Do you think that's failed as a concept?

[10:46 - 11:12] Fabian Seimer: I think the first or many companies are already moving away from this construct, definitely. Or grasp the area of digitization much narrower than we might do. Perhaps they are then only concerned with, let's say, digital, mobile customer processes or somehow special things like that and less with the overall picture. I think you can see that already.

[11:12 - 12:17] Alain Veuve: At Parashift, our mission is to say that we believe that IDP can fundamentally advance digital transformation, because in practically all processes that we digitize in large companies, there is still a document attached somewhere. Mostly one with a smaller volume. And that's where things start to stall a bit, and it's not so customer-friendly, because the documents are still here for a longer period of time, as it seems. Where do you think the journey is going with digital transformation in relation to IDP? I have a little bit of a feeling, maybe it's wishful thinking, as someone who is on the road here in this part, that IDP is a term that is becoming more and more important, but at the moment it's not yet so firmly anchored in people's minds, the awareness that every company has to have that. How do you feel about that? Do you feel like that's already there?

[12:17 - 13:23] Fabian Seimer: Certainly not everywhere yet. I still experience that today a really strong separation is made, so because, this is the scanning process, the OCR process, and then somewhere out the back then comes my business process, my effective business process. And I think this separation simply no longer makes sense today. I am in my business processes and in my business processes it now plays a role and you just have to think of it as a module or a building block that I add to my processes and then it works and then I can automate it. I believe that this change in thinking has not yet taken place, because today these companies also function in the traditional way. They may have a large scanning center and know the people in the logistics department behind it, and so on. I think there is still a lot to be done. But that's where we're at with the first topics where we talk about such things with the customers.

[13:23 - 14:08] Alain Veuve: Do you often come across companies that still generally maintain monolithic systems and have been taken a bit...I'm going to say something dangerous...hostage, maybe not intentionally by the vendors, it's more like...I often experience with large companies that have 300 developers, 150 of them are busy "keeping the legacy systems in the run somehow happy", I'm saying a bit nasty here. I have the wish perhaps that we come more to a service-oriented architecture, of course also because we live in this concept. How do you see that? Is there really that trend out there and is that something that CIOs should be doing or am I wrong about that?

[14:08 - 15:01] Fabian Seimer: Yes, definitely. I totally agree with you on that. That's also what I see. That is often really such a stumbling block for such digital processes are exactly the core systems or because on the front-end side customers invest very heavily. They want to differentiate themselves from the competition, they want to create a cool portal, a cool app that the customer likes to use. But then it just starts. What do I do with the data? How do I get it into my core system? And unfortunately, many core systems are still not really designed for this. In other words, the APIs that you need for this are really missing. Or rather, it is now the providers' turn to introduce this gradually. Unfortunately, it's like these initiatives are often blocked.

[15:01 - 15:20] Alain Veuve: Then you could say that 5 to 10 years ago it was maybe the mindset that had not brought digitization forward, so, you could have kind of, but you didn't really want to and today you want to but you can't....

[15:20 - 15:47] Fabian Seimer: Yeah, you can't is maybe too strong but it's certainly true what you're saying. The core systems just have to become much more open. In the past, the motto was a little bit, my core system just does everything and then there's just a little bit outside, but that's all not so relevant. And I think that's changing massively, because I can't get new customers with the core systems.

[15:47 - 16:03] Alain Veuve: Is it maybe also because of that...a thought comes to my mind right now...is it maybe also because of the meteoric rise of RPA to justify or to help justify that you can more easily overplay these shortcomings with RPA?

[16:03 - 16:21] Fabian Seimer: Yes, that is of course the means, if I don't have an API, if I don't have interfaces, then of course I can achieve a great deal with RPA and can also move forward here. Is certainly part of the reason for this topic, definitely.

[16:21 - 17:26] Alain Veuve: Exciting development, I think. Let's talk a little bit more about the mailroom. I've always thought, actually, every company has to have a mailroom solution, no matter how big or small it is. So slowly but surely I see in, actually so...I am after all industry stranger...yes, after 4 years one can say that perhaps no longer so but I am still one, if I look in our company, someone who has probably the least experience in the document industry. I'm always amazed at how little automation we see in the mailroom when we look at the broadness of the document cases. Do you think if we have better technologies there now, more out-of-the-box capabilities, that companies will adopt that quickly and move on there, or is the thinking like, "yeah we have paper," it's always about paper. In our world, maybe 40-50% is paper and the rest is digital documents that come in through other channels. Where do you see the development there?

[17:26 - 17:54] Fabian Seimer: Yes, I think definitely. So it's very simple. Today you always look at, what are my costs to automate this now and is it worth it? And in most cases it's just not worth it. But it's simply because these systems that we have today need quite a lot of maintenance in order to enable such automation in the first place. I think there will definitely be a change. Also in the mailroom solutions. Definitely.

[17:54 - 17:58] Alain Veuve: Do you think that's coming quickly or more like...?

[17:58 - 18:35] Fabian Seimer: That's a difficult question. So I think more the pressure will come from the front, from the business and less from the mailroom side, from the logistics side. That will come afterwards, I'm convinced. I think it has to come from the field, from the business, that's where the pressure will come from, that's where the need will come from, and maybe we'll also look at what we're doing in the mailroom. I can certainly imagine that. But I don't think we'll start on the mailroom side.

[18:35 - 18:38] Alain Veuve: Because there is no paying in the sense that everything is running.

[18:38 - 18:50] Fabian Seimer: Exactly, everything is running, and just, for the few documents that it then concerns, that's ultimately the reasoning, you also have to utilize the people that you employ there, those are the considerations.

[18:50 - 19:06] Alain Veuve: That means actually, if the costs or the efforts for the processing of small-volume document types fall below a certain threshold, then it is, how shall I say, the value of integrating them into the processes at all, that is the decisive factor, not the cost reduction.

[19:06 - 19:09] Fabian Seimer: Exactly, exactly, yes. That's what I would say.

[19:09 - 19:21] Alain Veuve: That's very exciting. Yes, in general, where do you think the journey is going with Intelligent Document Processing? Where will we be in 10 years?

[19:21 - 19:35] Fabian Seimer: That's difficult, that's difficult of course. I think in 10 years we will still see paper. So I'm not someone who says we're going to have everything only digital, I just don't think so, based on my experience.

[19:35 - 19:38] Alain Veuve: We heard that 10 years ago.

[19:38 - 20:11] Fabian Seimer: Exactly. And before that 10 years ago as well. So a paperless office that's really an old hat. I think we're going to see more of this process integration that we've already talked about. Maybe also networked between companies. At the moment, we always have the processes related to one company. But there are a bunch of processes that are always cross-functional per se, across multiple companies. And I think there's still a lot of potential to really put these processes together.

[20:11 - 20:44] Alain Veuve: Is that something that companies want? I actually experience in my daily conversations with customers a bit of, "yeah that's quite great, we can build a moat around our business model because there are inhibitions about sharing data and we might not want to." I never hear that so black on white or so clearly out, but I notice partly, the customers brake a bit and say, "I don't need to build an interface there at all, because then maybe the competitor can then take over my customers directly and things like that."

[20:44 - 20:51] Fabian Seimer: Yes, that is of course always such an issue, but nevertheless I believe that this is where the journey will go.

[20:51 - 20:53] Alain Veuve: So efficiency wins at the end of the day.

[20:53 - 21:10] Fabian Seimer: Efficiency wins. It goes to platforms, everywhere different platforms are coming to the market now already, where different partners are involved and have to share data with each other. I think that's the way of the future. And whoever is blocking that, yeah....

[21:10 - 21:12] Alain Veuve: Is that what we see in the insurance industry in Switzerland?

[21:12 - 21:17] Fabian Seimer: I'm also convinced in the insurance industry that this will play a role.

[21:17 - 21:28] Alain Veuve: Exciting, exciting. Yes, thank you very much Fabian for the learnings and insights and I look forward to seeing you again on another occasion.

[21:28 - 21:31] Fabian Seimer: My pleasure. Thanks Alain.

[21:31 - 21:37] Outro.