KNIME Business Hub - Basic Edition

I primarily use KNIME  for ETL, extracting data from different sources. I extract data from endpoints of Drupal  created for me by developers, then transfer this data into Oracle. After extracting, I create a model in Oracle with ETL, which is used by Power BI. Following this, I create a star schema of the data.

KNIME  is simple and allows for fast project development due to its reusability. I appreciate the ability to make improvements or modifications in existing workflows. Although I have not yet used the forecasting and customer profiling features, I find them promising. 

Another effective feature is the ability to use GET request objects to retrieve data from websites or APIs. This makes iterative steps easy to manage. It is more elastic and modern compared to SAP Data Services , allowing node creation and regrouping components or steps for reuse in different projects.

I have seen the potential to interact with Python, which is currently a bit limited. I am interested in the newer version, 5.4, when it becomes available. The machine learning and profileration aspects are fascinating and align with my academic background in statistics.

I have been working with KNIME for almost five years now.

Occasionally, when using the GET object, there might be issues due to the velocity of the lines or the IT system of the commission. Overall, stability is not a significant concern.

I have not encountered any scalability limitations with KNIME at the moment.

I contacted their technical support around five times. While they cannot always provide immediate answers, they are generally efficient and simplify tasks, especially in the initial phase of learning KNIME.

Neutral

I also worked with Power BI and BusinessObjects and have experience with typical data services access.

The initial setup was straightforward, taking between half an hour and an hour depending on the data entity.

I use SAP Data Services as well, but I find KNIME more elastic and modern.

I am impressed by the modularity and reusability in KNIME, especially the ability to make small adjustments to object configurations. I am interested in its interaction with Python and machine learning aspects. Also, I recommend KNIME to others as I face difficulty finding reasons not to. My overall rating for KNIME is between nine and ten.

I use KNIME  for my academic works.

KNIME  is more intuitive and easier to use, which is the principal advantage.

For graphics, the interface is a little confusing. So, this is a point that could be improved.

I have been using KNIME for six months.

I'd rate the solution seven out of ten.

We use the solution for data analytics and logic design.

The product is working fine with Oracle.

It is is written in Java. If they can output the Javascript, it will be much better. Also, it could be integrated with Visual Studio.

I have been using KNIME for three years.

20 users are using this solution. Scalability is quite easy, but handling many notes can become messy.

Most of the things is available in the community.

It's quite easy to setup.

I have a CSV reader. When I reset that CS reader, and It gave some error.

I have a CSV reader, and I encounter an error whenever I try to save. However, if I reset the CSV reader, I am able to save successfully. It’s a rare issue, but there's something wrong with the CSV reader. The error message doesn't provide a solution, only indicating a problem with the CSV reader.

I want to save the project but always face saving issues. If I reset the node, the saving works fine. The error message isn’t clear about what is wrong or how to fix it. I discovered on my own that resetting the CSV reader from green to yellow allows me to save the project. This issue is quite rare.

Last Friday, there was a widespread CrowdStrike issue, and I had to restart my computer. After restarting, I lost my entire project.

I recommend the solution.

Overall, I rate the solution a nine out of ten.

We used the product to prepare data for our team. I would prepare SQLs and check them in Oracle Developer, then create workflows in KNIME to manage and process the data, creating specific tables for modeling.

The product is a great alternative because it is not an open-source tool and offers simplicity, making it easier for our large team to use.

Sometimes, we needed more space to handle larger operations, especially since our machines had limited space and memory due to Kubernetes clusters. Breaking up SQLs was necessary to handle the data flow better.

I extensively used KNIME for about one year and at least two months.

The product is quite stable.

The platform is scalable. It is possible to configure the system to effectively manage memory and space requirements. 

I rate the scalability a seven out of ten. 

The community support is good, and plenty of shared knowledge is available.

We had licensing issues with other tools, but KNIME worked well as an alternative.

We integrated KNIME with Oracle, Apache, and other tools. It allowed us to pull data from various sources, such as Oracle, CSV, and Excel, into one consolidated table, which was very efficient.

Overall, I rate it an eight. It is a good tool, especially for our current requirements. However, there were limitations, such as space issues and occasional process slowdowns due to memory constraints. Despite these challenges, it is a solid product.

I recommend it to other professionals, particularly those who work with diverse datasets and require a flexible tool to manage data flows. It is user-friendly, especially for individuals with a background in Java or Python, as it allows for custom operations and automation, which I found very helpful in my experience.

Rather than specific use cases, I've used it in different sectors. Mostly education. For example, we've used it to predict the kinds of courses or degrees students should pursue based on their skill set and learning capability.

KNIME is very easy to handle and use. Anyone can use it, and it's easy to learn. You don't need a specific class. They're very good at model prediction. It has got everything.

From data analysis to model prediction and application, it's very good. I only use the free community edition, not the enterprise one. I feel KNIME is really good. I haven't tried any other tool or platform yet, but KNIME is pretty good.

The workflow is great. You drag and drop, and then you have the data explorer and charts that give results. 

The execution is also good – it's easy to identify where your model has gone wrong. It shows you the exact point of error within the workflow, so you don't have to execute the entire workflow to find it. For example, if your workflow has ten steps and the error is in the sixth step, it will show you the error at that step. You don't have to worry about the first five steps.

The Data Explorer is very good, and the charts are great too. The accuracy charts for different models, like decision tree, K3, Naive Bayes, are all very good. KNIME is great at reporting, whether it's structured or unstructured data. These are all very good features.

One disadvantage is the data load. KNIME has limitations. It doesn't handle large datasets or a high number of records well. I haven't tried more than 10,000 to 20,000 records because the model prediction doesn't come out well with more data. The Enterprise Edition might work better, but I've only used the Community Edition.

That's the only disadvantage I've encountered so far.  

I used it for four to five months, from the beginning of last year until January. 

It is stable. I would rate the stability a nine out of ten.

Scalability is also very good. Both the older and newer versions have improved. I would rate the scalability an eight out of ten.

I haven't needed it yet. I've been able to do everything myself.

I would rate my experience with the initial setup an eight out of ten, with ten being easy. 

I use the open-source version. I haven't explored the paid options yet.

Overall, I would rate it a five out of ten.