I mainly use KNIME Business Hub currently for data ETLs and then it meets with predictive analytics. Sometimes I utilize it for forecasting, but mostly it's predictive analytics.

I have utilized both model building relevance nodes in KNIME Business Hub, and there is the capability of deploying the model on the virtual PC so that everybody can utilize the model and send a query. Essentially, without an expensive decision engine, you can just deploy your model and utilize it, whether it's in an app, in something else, or on a website, it doesn't matter. You can just ask a question and the predictive model runs and you can get the results.

Collection of company-wide information is one of the main benefits that KNIME Business Hub provides to the end users; all the intellectual property that has been developed in a central location is critical. When somebody leaves the company and another one comes in, you can have all the information on KNIME Business Hub about what's working so far, what's deployed, and other relevant details. That actually provides a lot of information, so instead of doing something on your own computer, it's all centralized. It actually increases the general maturity of the company, including not only the user but the general maturity of the company and the line of businesses. All of the workflows are stored; I know who utilizes it, who built it, when, and how frequently it's used. A lot of insights are important, but in general, I have a lot of intellectual property built-in so I can keep track of it. It's very similar to GitLab, but this is for analytics and for the company itself.

On the computer vision part within KNIME Business Hub, it's not that capable, unfortunately. I utilize some of the existing nodes, but other than that, if I am handling classic tabular data, it's fine. If I'm handling unstructured data sets, I can push it to large language models and get some results back, and that's also fine. However, when it comes to visual analysis or vocal analysis, that lags a bit.

Integration capabilities of KNIME are almost simple. I have integrated with enterprise resource planning systems, SAP, and all that. With SAP it is working well, but it could have been better.

I would describe KNIME Decision Hub as somewhat helpful in making data-driven decisions more efficient. It could have been a scalable decisioning as a service at the back end, but it's not working that way. I can deploy a model and get some results by sending some queries, but if I happen to utilize it for a more scaled-up business such as a bank or telco, that is a different situation. Then, I start requiring hundreds, thousands of dollars worth of licensed software, and I wish it would be capable of scaling up and down as needed.

Visual analytics is the main point for improvement for KNIME Business Hub. Computer vision is the most important because now there is a new age of large language models and visual language models. The visual language models can turn an image or a video into text, and you can utilize it as if it's a very capable computer vision model. It tracks all the segments and all the labels on the images or videos, which means that if I can interact with these through KNIME Business Hub, then I can build very sophisticated analytics end-to-end. Right now, that's not that much possible, but I wish it's going to be in the near future.

If we talk about functionality, I would like to see most of the classic independent large language models and visual language models integrated into the next version of KNIME Business Hub. There are new multimodal capabilities in it, and I have to go grab a model from the open-weights structure, implement it somewhere, and send some queries. I wish that most of that would have been built-in. Databricks, for example, built models embedded into their software structure so that I don't need to go to a third-party; I can run some models to enrich my data. This includes enriching images into tabular data sets or converting voice into tabular data sets. Many enterprises cannot just push the existing data outbound and get some results back since a lot of that consists of internal data sets.

I have been working with KNIME Business Hub since 2021.

Most of the cases that I utilize from KNIME Business Hub require data from databases with semi-structured, almost clean data sets. However, more and more required data starts coming from Internet of Things devices, and these are streaming data sets, not static data sets living in databases. I need event listeners that listen to something, on-the-fly score it, put it into context, and send it to a large language model. Complex event processing is becoming much more of a requirement, and in that case, I need milliseconds of performance. KNIME Business Hub lags there a little bit, but it's improving.

First, I deployed KNIME Business Hub on-premises, such as on my machines, because it is low-weight. I can immediately install it onto my personal computer or even my tablet, run some data on it, and get some insights immediately. I don't need a huge footprint or an expensive hardware-plus-software combination. That's a good aspect. But at some point, if I want to excel in a company with one hundred users or fifty users, it has to be on either my own servers or a private cloud.

KNIME Business Hub can also be deployed on Amazon Web Services, Azure, or Google Cloud Platform based on the client's requirements. Amazon Web Services is the best option, but I haven't seen anything I can purchase from Amazon Web Services's marketplace. That would be a good benefit for any client to onboard immediately. Just click it down from the marketplace, and if I have corporate credits, I could utilize it there so that suddenly everybody at a bank, for example, owns a KNIME Business Hub license. That would be awesome.

Pricing for KNIME Business Hub is much cheaper than the competitors. It's a reasonable amount of money for the product.

I am already advising many companies to start machine learning and artificial intelligence integration with KNIME Business Hub because a lot of the large language models need to be integrated with classic basic machine learning to turn it into something called composable artificial intelligence. I generate a lot of data from text and video, push it into a machine learning model, push it to a forecast, get the results, and then export it with a large language model, talking as a normal person would. This is the composability part of it. If you want to do composable artificial intelligence and get return on investment fast, go with KNIME Business Hub. Don't go to software as a service, Databricks, or IBM Watson. These tools are both expensive and very capable but will take months to build something and get positive return on investment, leading to months in negative return on investment. With KNIME Business Hub, you can immediately grab some data and see positive return on investment. They have single-user free licenses, so I can start now on my Mac, throw some results, and do a proof of concept or proof of value for upper management in just a few weeks. Then I purchase KNIME Business Hub, and the company begins benefiting, leading to immediate positive return on investment. That is the difference that many don't understand. I would rate this product nine out of ten.

I am currently using KNIME Business Hub. In my experience, using KNIME Business Hub as a unified platform for developing advanced analytics and artificial intelligence solutions enables distributed processing of large-scale data through Spark. Implementation of modern lakehouse architectures that integrate data engineering, data science, and analytics within a single environment enhances scalability, model versioning, and team collaboration. Currently, I use KNIME Business Hub to build data pipelines, train models, and deploy analytical solutions into production environments.

I am also using other tools because my company has many clients and our clients have different tools. We need to construct the analytical solutions in these tools. For example, I am using Python because in Python we construct the statistical and analytical models. Python is the primary language for developing advanced analytics and artificial intelligence solutions, including machine learning, deep learning, and large-scale data processing. My company has strong experience with different libraries, such as Pandas, NumPy, Scikit-learn, and TensorFlow. For our clients, we need to build, validate, and optimize predictive models. My team is multidisciplinary, and we integrate solutions into production environments through APIs, process automation, and end-to-end analytical pipelines, ensuring scalability and maintainability of the models. I always use Python as well. However, I use KNIME Business Hub in the same way because KNIME Business Hub is very important for constructing advanced analytical models. KNIME Business Hub now has many nodes to use for big data, data quality, data governance, and advanced analytics. We use KNIME Business Hub as well. It depends on the client because we always try to analyze what tool our client has, and then we try to use this tool. KNIME Business Hub is another tool that we now use, and we use the Python nodes as well for advanced analytics. In data governance, we try to use KNIME Business Hub to construct the data quality rules and other analysis. For example, to assess and understand the maturity of the companies, we sometimes use KNIME Business Hub. I use different tools, but sometimes KNIME Business Hub, and other times Python and KNIME Business Hub are different tools. I also use Amazon Web Service and Azure.

My experience using KNIME Business Hub for the development of advanced analytics and machine learning solutions leverages a wide range of nodes across data preparation, modeling, and deployment stages. I always try to use specific nodes because we always try to use the CRISP-DM methodology, so we need to always do data preparation and transformation for advanced analytics solutions. Key nodes and components used include data preparation and transformation nodes such as File Reader, Row Filter, Column Filter, Missing Value, String Manipulation, Math Formula, Joiner, GroupBy, Pivoting, and Rule Engine. I use nodes for feature engineering, such as Normalizer, One to Many, Binner, Lag Column, and Feature Selection Loop, and other nodes for machine learning and AI. For example, Partitioning, Decision Tree Learner, Predictor, and Random Forest Learner are all models that KNIME Business Hub has, and we use them for our models. Sometimes, I always try to use the Python and R nodes because there I can program the code as well. For model evaluation, I use other nodes, such as Scorer, Confusion Matrix, and Numeric Scorer. I love KNIME Business Hub because I can construct workflow automation and deployment. For me, it is very clear to understand the process for constructing analytical and advanced statistical models. It is good for me to use KNIME Business Hub for that. I use KNIME Business Hub end-to-end, from data preparation and feature engineering to machine learning, model evaluation, and workflow automation, integrating Python and R when more advanced modeling is required. I always try to use KNIME Business Hub.

It is very important that I have the workflow automation integrated with Python nodes, for example, and I can construct our main code to construct the solutions. For us, it is very important to have the workflow automation. In KNIME Business Hub, it is possible because we have the end-to-end approach to the models. We have, for example, some nodes for data preparation, and other nodes for feature engineering, and other nodes for machine learning and model evaluation, for example. We have only one workflow with all the nodes and all the processes. For us, this is an important impact because, for example, we have to construct segmentation models for our customers, and we define a frequency to run the models. For example, we need to run the cluster segmentation around each month. We have the automation of the workflow and we need only to put a run in a button and the process runs. For us, this is an important impact because the time to obtain the results is very quick.

Sometimes it is a little bit difficult to use some nodes when we have many large-scale data, for example, CSV files with a large amount of data. It is sometimes difficult to try to import the data in KNIME Business Hub nodes because I think that some features that are in the CSV in text, for example, large text, is difficult for KNIME Business Hub to import these fields. I don't know why, but it is very difficult. We need to try to use different nodes for importing the data, such as File Reader and CSV Reader. However, I think that it is always the features that have much text, it is difficult for KNIME Business Hub to understand and import this information. I don't know why, or maybe I don't know if we don't know what the better option is to configure the node to import all the CSV or the data set. However, we have always had this problem. In some nodes, sometimes it is the same because sometimes, for example, I have a CSV and in my CSV, I have a feature that is, for example, a date. When I import this data set in the File Reader node, I have problems with this field because it is a date, but the problem is that it imports it as text, for example. We try to use their nodes that convert text to date, but sometimes it is difficult, and it is not immediate to transform the text into a date. So we needed to convert the text into a date in the CSV, and then import it again in the KNIME Business Hub node and try to have a good read of this field. I know that KNIME Business Hub has some nodes to convert text to date and others, but sometimes it is difficult to use these nodes. I don't know why. Maybe it needs a specific format for the date and we need to transform our feature in this option. So sometimes it is a large process to convert these features. However, sometimes we need to investigate and search for other nodes, and try with other nodes to import these cases.

I started with KNIME Business Hub around fifteen years ago.

For me, it is great. I think that sometimes we have some missing problems in some nodes when we are constructing the statistical models, but we always try to visit the forum for KNIME Business Hub and then we try to resolve the problem. However, I think that for now, I need to come back again to Germany to make another training because I saw that KNIME Business Hub now has many new nodes and I need to explore the new nodes and try to use more. For now, KNIME Business Hub is excellent for me and for our team.

We are a partner from KNIME Business Hub at this moment and I made different certifications in Germany, in Berlin, with KNIME Business Hub about machine learning nodes. I think that was around 2016. In 2018, we made two certifications with KNIME Business Hub.

For now, we always try to use KNIME Business Hub to integrate with Power BI because we use Power BI to present the results and the visualization for the models. In KNIME Business Hub, I try to use some graphics, but for our internal analysis. For our clients, we use Power BI to present the results for the models.

I think that KNIME Business Hub is very robust and is a leading solution for analytics and advanced analytics. I think that now we have many nodes to construct the analytical models in the big data nodes and to process structured data. This is important because it is very easy to use the nodes in KNIME Business Hub in these cases. For example, in Python, it is a little bit complex to construct the code. In KNIME Business Hub, we have the end-to-end approach to the workflow, the complete workflow to resolve the process for the model. This is very good to have good results and quick results for advanced solutions, for analytics and for artificial intelligence. I think that I prefer KNIME Business Hub to Python, for example.

I think that the price is good. I think that a good option is to analyze, for example, the cost for Amazon Web Service, AI components of Azure and Amazon, and try to compare to KNIME Business Hub, and I think that it is a good price. However, always in our solutions, we need to make a good calculation for all the solutions because we have many solutions, and because all our clients don't have KNIME Business Hub. Sometimes we use KNIME Business Hub for our internal development of the analytical models. However, sometimes our clients have KNIME Business Hub, so it is perfect because we can construct the models there. When our clients don't have KNIME Business Hub, we need to use other tools because sometimes our clients tell us that they need us to construct the model only in their tool, for example, Amazon Web Service or in Python, so we need to construct there. Because sometimes they don't know about KNIME Business Hub and they want to use the tools that they have. However, I think that it is comfortable to use KNIME Business Hub for our clients. They like it very much because it is very easy and now it is very robust for statistical and advanced analytical solutions. My overall rating for KNIME Business Hub is eight out of ten.