What do you think of Aspiration Bank

Interview story by Mohammed Brueckner about cloud computing

Mohammed Brueckner is a technologist, CRM and e-commerce expert and is interested in cloud computing. He loves to talk about platform economies and has used our bimanu cloud as an opportunity to conduct an interview story with us regarding data integration and analysis. We published the result in our blog post. At this point, many thanks to Mohammed for the interesting exchange.

introduction

Every company must gain further knowledge in the future in order to maintain or expand its competitive advantages. This is based on the data labeled the new oil, although the comparison lags a bit as the data is not consumed.

It is no longer sufficient to just look at the company data, but also all external data, such as marketing information that is determined via campaigns or website visits to see how high the actual conversion is, must also be viewed in the common context. It becomes even more difficult when unstructured data such as streaming data (Twitter & Co.) play a role. This data must be related to the classic business data using different systems and technologies.

The way there is by no means easy and represents a great challenge for all companies. It becomes even more complicated, especially for small and medium-sized companies that want to consolidate their distributed data areas and thus make them analyzable. You can not afford the Big Data, Data Science and Business Intelligence units or these experts are currently very difficult to get on the market. Larger companies recognized the need years ago and created the appropriate organizations and, if necessary, can also involve external service providers to set up tailor-made analysis platforms, data lakes or business intelligence systems.

But that's not all, such complex projects have a high investment volume in terms of project implementation, procurement of hardware and software licenses and the use of external service providers and internal resources. If you've ever been involved in an endeavor like this, you know that implementing a process like this can cost millions of euros. In addition, the follow-up costs for operation must be taken into account.

It is a good thing that in times of the democratization of the most advanced instruments and services we have a multitude of options and are not dependent on introducing the large standard solutions.

bimanu has set itself the task of supporting and accompanying companies of all sizes and in particular medium-sized companies with an all-in-one platform for data integration and analysis that differs significantly from the previous SaaS offers and does not rely on a single proprietary all-or-nothing -Nothing offer relies on the combination of best of breed tools.

I'm Mohammed Brückner and this is an interview with Swen Göllner, founder of bimanu and an experienced BI veteran.

"Swen, could you tell me a little bit about you and bimanu?"

“Very much Mohammed, first of all thank you very much for the interview and the interest in our solution. Briefly about myself - I am married, the father of 3 kids aged 2, 6 and 11 years and have been self-employed with bimanu in the field of business intelligence for 3 years. I have been dealing with the topic of data & analysis for over 15 years, originally started in the application company, later gained further business intelligence experience in a consulting company and have been working as a business intelligence consultant in the banking and insurance sector in recent years.

As a part-time job, I completed two courses in business informatics and general management. In the end, the MBA was also the decision to choose the path to self-employment, as the Entrepreneurship lecture series showed the ways which freedom is possible with a look outside the box. In addition, I am very curious and question things and do not always try to follow the standard route. In companies there are clear hierarchical structures, processes but also corporate cultures that have to be taken into account and which limit me as a person quite a bit.

Before 2016, together with my current business partner Michael Jungschläger, I already had the idea of ​​a simple evaluation platform only for end users who can analyze your data without IT knowledge and who do not have to think in advance how I can get this data in the correct shape together. We carried out the implementation part-time at the beginning of 2015. Initial discussions with software companies, business angels or incubators were sobering.

So we made the decision to terminate our secure employment and to found bimanu GmbH and that in a phase where others thought how could you take such a risk?

Right at the beginning we were also able to get Dr. Jon Nedelmann gained experts as an experienced data scientist and the development of the bimanu cloud was started together.

Nevertheless, we were well prepared in terms of financing and the market situation, because we started with what we are good at - the business intelligence service and, at the same time, our software idea bimanu Cloud was implemented with the bimanu team. We have been on the market for 3 years now, our team has already grown to 7 people and the bimanu Cloud has been on the market since the beginning of 2019.

At this point I would like to thank Michael & Jon in particular, but also the rest of the bimanu team, for the 3 great years so far. "

"What is the value proposition behind bimanu, in your own words?"

“For us, in addition to high quality standards, there is also reliability. At that time, we even listed reliability as a competitive advantage in our business plan. In concrete terms, we stand by our word and only offer what we can actually do. Our advantage is that we got to know both sides very intensively, the application companies in the IT & specialist area and the consultant side.

We understand both worlds and it is precisely this requirement that we have of ourselves in classic services that we have transferred to the bimanu cloud. It is important to us to always come from the requirements side, i.e. what do the end consumers want to achieve with the data and what is their purpose behind it? In return, IT tends to focus so much on technology and technical implementation instead of keeping an eye on the specialist area. Then it can happen that the consumers use the data and analysis platforms, but the further refinement takes place in the departments and this creates a new shadow IT - different interpretations of the data or an inconsistent database are the result.

We don't want to discuss the technology, it is important, but only as a supportive measure, in the end it is all about the customer. In the last few years the possibilities have been increasing, but the technologies are becoming more and more difficult to control - also for the IT ’ler. Consequence: Specialists are needed who are not fully available. Worst of all, however, are the so-called buzzwords IoT, AI or BigData. These are taken up by all parties concerned and named as a salutary solution for digitization. This is definitely not the case, rather a solid database must be built up in the first step, otherwise the data scientist colleague is left on dry land or is responsible for the data integration himself. Quality-assured - and very importantly - historical data is also required for the topic of machine learning. A consideration of only 2 months is therefore not sufficient to enable learning of the program. "

"What are the biggest challenges you see today for medium-sized companies trying to improve their game in terms of reporting and advanced analytics?"

“Medium-sized companies have long recognized that business intelligence can be an important tool for corporate management. We've been able to speak to some medium-sized companies over the past few months and have identified the same problem for all of them.

The relevant information for corporate management is usually available in distributed systems and is difficult to combine. As a rule, attempts are made to prepare and evaluate this data manually in various evaluation tools such as Qlik, PowerBI or Excel. The consequence is recurring effort in the analysis and reporting. In addition, the data must be organized from the source to the analysis itself and the risk. that the data quality suffers is high.

Many are now also positive about the topic of Cloud BI, which was different a few years ago. Barc also came to this conclusion in its study “Cloud BI continues to advance” - https://barc.de/news/cloud-bi-weiter-auf-dem-vormarsch.

The advantages from a company perspective are obvious. Fast implementation compared to on-premise solutions. The cost reduction: There are no license, maintenance or hardware costs, no additional employees are required and only what is actually used is paid for. No special expertise is required, as only a specific application scenario is used. This helps companies that are unable to create their own data integration and analysis platforms due to their limited resources.

But not only the technical requirements are to be named as a challenge, the software provider market is also intransparent and does not exactly make it easy for the decision-makers. Of course, Gartner Magic Quadrants can be used as comparisons or you can orientate yourself on the well-known software manufacturers such as SAP, Qlik & Co. who promise the self-service approach with their marketing. Don't get me wrong, these are good software applications, but they are not enough. The software systems only form the basis and the end user is responsible for his data from the source to the analysis. "

"In your opinion, how have compliance requirements and legal obligations such as GDPR affected the way companies gain a better understanding of customers?"

“Interesting question - in fact, the way data is handled has changed. In the past, it was a matter of course to prepare the data quickly using Excel and make it available by email, or personal data could also be viewed in the test environment. This is definitely no longer the case due to corporate compliance. The efforts are enormous in the area of ​​test data management.

The anonymization of personal data must be guaranteed, but the data integration tools used do not usually offer such a functionality. For everyone involved, this effort in the projects should not be underestimated. In the banking sector in particular, the requirements (e.g. BCBS 239 - principles for the effective aggregation of risk data and reporting) are extremely high and most productive data warehouse systems do not meet the requirements and must be adjusted afterwards with great effort, if this is so easily possible . Comparable to the conversion of a cellar in an existing property. "

“Nice metaphor! Let's assume that I am a medium-sized company and would like to use agile process models, e.g. to create a minimum viable product (MVP) with the aim of awakening customer and market needs. What would be a sensible approach on your part or what did you see in your work? "

“The feedback from customers is extremely helpful in developing a needs-based product, with the advantage that the product does not have to be perfect from the start. Our solution is based on the Scrum method - a process model in the project management environment - with the help of which we carry out the requirements elicitation incrementally and iteratively on the one hand with the help of user stories, on the other hand our data model is aligned according to this agile approach. We can present the interim results in a few days and the customer can influence and design his data model in such a way that he achieves the greatest possible benefit. Another advantage is further development or adaptation. New interfaces or new data fields are no problem for us. Our model is agile enough to include new developments without affecting the existing model, the advantage - the test effort for the previous model is eliminated and the test scenarios only affect the new development. "

"Which analysis solutions are supported by the bimanu cloud? And which solution would you recommend to which company and for what reason? "

“The bimanu Cloud is an all-in-one solution and combines all steps from integration to the data model in one platform.

In addition, our self-developed TriData model can, as already mentioned above, record company data from all possible areas and link this data with marketing information, but joint evaluations with IoT streaming data are also possible. The interesting thing about it is that, depending on the analysis requirements, we enable an inventory or inventory movement view based on these 3 data areas, i.e. either a point in time or a period perspective.

As part of the bimanu starter package, the exact analysis requirements are determined with the help of user stories and the required data model is automatically generated on this basis. In this way we ensure the exact requirements. The recommendation is always based on the customer's requirements, so no general recommendation can be derived. However, it has proven useful to start with a small evaluation range and then add new information in further iterations. The stored data model is designed precisely for this cyclical growth.

In the first step, each customer then receives a dashboard in the modified bimanu Tableau Online environment and thus receives a standardized 360-degree company view of the available data. "

“So I assume that you covered the tooling and integration support, that is, all the technical details. What would you advise the customers in terms of organization and culture when they set out to become "more data driven"? "

“That is exactly what we have achieved with our solution. The customer receives an all-in-one platform including operation and support from our bimanu team. But what exactly does it mean for a company to be data-driven? For example, the following requirements must be met:

The company has to collect data - not just any data, but the correct information relevant to the question. And these must meet the following requirements: prompt, accurate, quality assured and trustworthy.

The data must be accessible and analyzable. Exact, timely and relevant data is not enough in the first step to be considered data-driven. The data must be available in a uniform form, which can be linked to other company data if necessary. So simply relate the structured data to semi-structured data (BigData, IoT, Streaming). Is it that easy for today's companies? I do not think so.

Companies must agree to a permitted data exchange culture within the organization so that data can be linked, e.g. the combination of the customer's clickstream with their transaction history. This works best with a quality-assured starting point.

Another important prerequisite is the correct selection of an analysis tool for querying the data. All dashboards and analytics require filtering, grouping, and aggregating data to reduce the large amounts of raw data into a smaller set of high-level numbers that help us understand exactly what is happening in an organization. In addition, a central storage location for the objects is required, where they can be accessed either on the desktop or mobile and at any time without delay. "

“OK, what do you think - where are business intelligence and data analytics in general going to take a look inside the crystal ball? And please tell us which investment is really sensible from your point of view and which is not. "

“The trend can already be seen. More and more tools show a self-services approach, i.e. the analyst gets enough help with data analysis and interpretation of his data. The market research results from Gartner, Lünedonk and Barc also show that the business intelligence market continues to grow.

In the future, the analytics tools will increasingly implement the self-services approach towards data scientists. But here, too, the use of machine learning will make more and more automatisms available, so that in the end the analyst must be able to evaluate the result of the algorithm.In the integration environment, some ETL tools have a high degree of automation, in particular through the Data Vault data modeling method, more and more data integration processes can be further automated. Machine learning uncovered even more potential - what if the program automatically generated the correct data model from the metadata? There are the first startups that have implemented their first environments in this way.

Investments in the classic sense in the form of on-premise products are no longer up-to-date, i.e. the setting up of server infrastructures, manual installations and so-called customizing can already be replaced by appropriate SaaS providers. Companies should use the micro approach for their specific operational requirements, there will be no software for all areas. Companies should rather fall back on specialized solutions that fulfill a specific purpose and can be easily networked for later comprehensive evaluations.

This can be a wise investment. Cloud approaches can offer added value here, as these systems can be evaluated quickly and at low cost. We are also far from the end of development with our bimanu cloud solution - voice control for analysis or machine learning for automatic error handling are just a few points that we want to implement in the future. "

“Thank you, Swen! I guess it is clear that bimanu has the tools to help companies that are not yet sure how to deal with the challenges of advanced data analysis. As you rightly say, technology is a dimension that needs to be mastered, but it is by no means the only one. Interested parties can find out more at bimanu.de. "