Community-driven innovation is constantly opening doors to the very latest analytic techniques. It’s one of the reasons KNIME believes in openness and the power of the community! Members of the vibrant KNIME community dedicate themselves to sharing their knowledge, developing new functionalities, and offering mentorship.
We got to know two members of the community in a panel discussion at the Community Awards for KNIME Excellence (CAKE) which celebrated special community Contributors of the Month.
At the CAKE Talk, Raffaello Barri, data science consultant at BIP, Italy, and Arjen Peters, Aircraft Data Consultant & Support Engineer at Exsyn Aviation Solutions, The Netherlands, share their journey with KNIME, delve into the potential of a low-code tool, and offer their perspectives on the future of data science.
Meet our guests
Raffaello is appreciated on the KNIME Forum for his valuable contributions answering all kinds of questions by beginners on using KNIME. He was awarded the COTM award in December 2022 together with 3 others, for his technically advanced and end-user centered solutions to the Just KNIME It! challenges.
When you visit Arjen Peters’ profile on KNIME Forum you can see from the stats how active he is in providing technical support to the community in general, and to intermediate KNIME users in particular. To honor all his work, he was awarded KNIME COTM in June 2023.
Read a transcript of the interview below.
***
Roberto: Let’s briefly introduce yourself, your professional self, and how you work with data.
Raffaello: I’m Raffaello, I’m 28 and I’m based in Milan. I work in management and I originally started my data journey using Excel. I had my ups and downs and then I moved to KNIME. The switch was very smooth and, once I was comfortable with the software, it saved me a lot of time.
Arjen: I’m an Aircraft Data consultant at EXSYN Aviation Solutions and KNIME is the platform we use the most. Our clients are usually airlines and maintenance organizations, and one of our main tasks is to help them migrate their data. Put simply, we extract data from a source system, we transform it according to business and technical roles, and we create an output that a new system requires.
Roberto: How do you think that KNIME can be of help to anyone working with data, any data analyst/engineer/scientist independent of their specialization, job profile, and educational background?
Arjen: I believe KNIME can be a very useful tool because of its versatility. You can use it to perform very basic operations but it can also get as complex as you like. I really like that you can create your own extensions or use the one created by other members of the community. In other words, you can get as creative as you like.
Raffaello: Arjen has made a good point! I think KNIME allows you to focus on the things that really matter and when I say so, it comes from personal experience. Back in university, I took a deep learning class that, coming from a management background, was something pretty new to me. On top of studying the theory behind machine learning, I was also required to learn how to code in Python. I had to spend a lot of time learning how to program, instead of focusing on the design of machine learning algorithms. If I had had KNIME, I could have focused on the results, minimizing the time spent on secondary aspects.
Another thing I appreciate is that with KNIME you can hide the complexity of the tasks you have performed. If your colleagues are not familiar with modeling or crunching numbers, you can make them play only with a few parameters. It’s a win-win situation for me as a developer, because I let my users operate in a controlled and limited environment, and for the users who have the possibility of getting the information they need in an easy way.
Roberto: Let’s talk about your personal experience. How did you come across KNIME Analytics Platform? And what was your impression at the time? When was the wow effect, the turning point?
Arjen: In my company, we were in many cases reinventing the wheel for things that could be automated. The projects we do are usually pretty similar to one another, so we were looking for standardization and that’s why we chose KNIME. We were all used to ETL tools and we transitioned smoothly into using KNIME. The main reason why we chose it was the ability to create our own components.
Raffaello: My former boss introduced me to the low-code approach and suggested KNIME to start off in this unexplored territory, but I had no time to look into it, since I was working on my Excel, PowerPoint and PowerBI skills. After some months, I came back from a vacation and still had some days off work so I completed the basic and advanced proficiency courses with KNIME. My boss explained to me that those were some important yet initial steps, so I pushed myself to use KNIME every day at work to really unlock the full potential of it. After that, I was like: “Excel who?”.
Roberto: How did KNIME help you with your work? Did it make it faster, more accurate, more agile, or what else?
Arjen: For me, reusability and automation were key. We have created many standard workflows and even our library of components. Whenever a customer contacts us, we most likely already have a workflow ready for them to use by just changing the input data. It takes us only one hour to have a solution up and running.
Raffaello: If I had to pick a word I'd say “independent”. When I am faced with challenges, what I truly appreciate is the ability to tackle them independently. Regardless of the level of difficulty, I find satisfaction in being able to find solutions on my own without having to wait for someone else to provide the answers. Yet, as a beginner you usually only know the basics of the software. Therefore, whenever you have issues you can leverage the community support on the KNIME Forum, ask questions there or look for the solution by browsing already resolved tickets. In this way, the journey to data analytics independence is much faster thanks to KNIME’s vibrant community.
Roberto: Can you name for us the KNIME feature or node you could not do without?
Arjen: In our case, it’s the Column Expressions node, since we have to do a lot of calculations that are based on multiple conditions. It’s easy to use because you can select the function you need by just clicking on it. You will find this node in all of our workflows: It’s very powerful!
Raffaello: I would also pick the Column Expressions node. It can do many operations that would otherwise require multiple nodes. So when I need to get a job done quickly, it’s really my go-to option.
Roberto: Raffaello, you learned KNIME Analytics Platform roughly 1.5 years ago. Is it true that KNIME is easy to learn? How long did it take you to learn KNIME Analytics Platform?
Raffaello: I’m glad to say that the learning period never ends. The platform has so many features that there is always something new to learn. Citing the Pareto Law, once you acquire 20% of knowledge, you can solve 80% of problems. To get to that 20%, I would say it took me a couple of months. I was lucky because my boss pushed me to use KNIME as much as possible so I was able to practice a lot. In Italian, we say batti il ferro finché è caldo [strike the iron while it's hot], this is to say that when you encounter something extremely useful and cool, you should dive in 100% to challenge yourself and learn more about it.
Roberto: Did the Just KNIME It! challenges help you familiarize with parts of the software you did not know?
Raffaello: Yes, absolutely. Many times when I was stuck, I said to myself: “Oh, I know where I saw the solution to this before”, so I would go back to the list of Just KNIME It! challenges and look for the one I needed. These challenges are a great way to familiarize yourself with problems that are more “niche”, and at the same time they offer solutions that can be repurposed elsewhere.
Roberto: You both are very active on the KNIME Forum. What is your username on the KNIME Forum and which kind of questions do you usually answer? Do you have some kind of specialty in answering Forum questions? What’s the most difficult question you answered?
Arjen: My handle is @ArjenEX and I help mostly with core features for data manipulation. Another thing that many people struggle with is recursive looping so I try to help with that too.
The hardest problem I solved was about how to handle API responses in a recursive loop. It took me quite a while to figure it out, and I talk about it in more detail in the chapter I wrote for the book “Best of KNIME: The COTM Collection - Season 3”.
Raffaello: You can find me as @Lelloba. Anytime I see a question I know the answer to, I write it down. If I can help, I am always happy to do so. I see it as giving back to the community all the guidance I received when I was still inexperienced.
Once I was trying to help a user with a question that was not too technical. The problem though was that we could not really understand each other in writing, so we jumped on a Zoom call. I went into the call without a clear idea of what the problem was, or whether I was even able to solve it. After one hour and a half, though, we figured it out.
Roberto: The world of data science is constantly evolving. Every few years or even every few months there is a new disruptive improvement in the field. The talk of the day is the new Large Language Models, the new AI models. Do you use the new KNIME nodes for AI in your work?
Arjen: Aviation is a very regulated industry and at the moment we all know the issues between AI and regulations. We definitely use some NLP nodes sometimes and we have a lot of recent graduates working in that field. Although we are still in the development stage, our focus involves training a model using data extracted from knowledge-based repositories. The ultimate goal is to create an interface where users can simply prompt something like “I'm encountering this issue, can you help me solve it?” and receive a detailed response instantly.
Raffaello: I played with these extensions just because I was curious about it. However, the data privacy concerns around LLMs requires careful evaluation by businesses so in my daily work life I do not really use them.
Roberto: Thanks, Arjen and Raffaello, for the great conversation and for sharing your thoughts on KNIME, automation, learning, support and AI.