My main use case for Dataiku involves ETL pipelines, mainly for data analysis, and I majorly use SQL queries for that.

For ETL pipelines and data analysis, I had to create the output by combining a few datasets and then running SQL queries, applying filters, joining the tables, and so on; so I used Dataiku for that.

Regarding my main use case with Dataiku, I primarily use it for analysis only, and the visual recipes of Dataiku and the SQL query are enough for that. No challenges have occurred so far, but the only challenge is that Dataiku gets slow sometimes and lags a lot.

The best features Dataiku offers in my experience are its visual recipes, which are very easy to use for analysis.

The visual recipes are easy and useful for my analysis because the Sync recipe is very useful if I want to download a table from the cloud into the Dataiku database and schema. Other recipes such as the Prepare recipe are also very useful since you don't have to write code; it's all visual and very easy to use. Recipes such as Stack are also very useful as you don't have to write full SQL code for it, allowing you to speed up the process.

Dataiku has positively impacted my organization since we use it majorly for our day-to-day work, and it is very helpful in creating and managing ETL pipelines to create a project flow, making it easy to go back to any step and then make edits if some changes occur.

I have no suggestions for improvements because it's all good; it just sometimes lags a lot, and I don't know if the server is full or what, but it sometimes takes a lot of time while loading and refreshing the page.

No additional thoughts on improvements have come to mind, but the performance can be more optimized to reduce the waiting time. Dataiku is down a lot of times, and we have to wait for sometimes five, ten, or fifteen minutes, after which it gets working again, and during those times, we are unable to get our work done.

I have been using Dataiku for four years, so my experience with it is quite extensive.

Dataiku is stable for most of the time, but for around ten percent of the day, it is usually down, and we are unable to do work on it.

Dataiku's scalability is good.

I have never needed the requirement for customer support from Dataiku.

I have been using Dataiku for the last four years, and I have not used any other solution besides Dataiku.

It is a good return on investment since it helps save a lot of time, and it's easy for my teammates to work cross-functionally on the same project.

I did not evaluate other options before choosing Dataiku because it was all managed by my organization, so I had to use Dataiku only.

My advice for others looking into using Dataiku is that it's a good software, and I would suggest them to keep using it since it's a very good tool for data analysis uses.

I have no additional thoughts about Dataiku; it's all very good for the use cases, but if the performance can be improved to be more stable with lesser lags, it would be much better. I would rate my overall experience with Dataiku an 8 out of 10.

My main use of Dataiku is especially data preparation. I use Dataiku a lot for preparing my data, in particular processing and transforming my datasets by using recipes, creating recipes, and especially what I really value is being able to reuse the recipes already created for preparation on another dataset.

I was preparing for my Core Dataiku certificate, and all of the modules were focused on data preparation. I load the data into Dataiku, then I use the recipes and tools to add columns, unpivot columns, delete, and transpose columns so I can format them. Then I create groups of recipes and I also reuse them by importing another dataset into Dataiku, which gives me the ability to save time. I don't have to redo all the previous processes since I already have a recipe for data preparation that I can reuse.

After data preparation, I had the opportunity to carry out an end-to-end project with Dataiku. This involved first the data preparation and then I went on to set up a model to predict a stock index. I used the machine learning models for this project.

Let's suppose I have datasets that I use every time, and each time I'm going to check the data formats to format a certain number of columns, for example dates, to see if they are in date format or not, delete certain columns, rename certain columns, transform the data, and clean them. If I've done all these steps once and I manage to put all these recipes into a group, next time it's an enormous time saver not to have to repeat these steps one by one, but to use directly the recipe or the group of recipes created.

I want to emphasize again the recipes and how we can reuse them with Dataiku. In most data projects, data preparation takes a huge amount of time for professionals, and sometimes unfortunately we repeat the same tasks. Dataiku really brings a solution to this in that we can create groups of recipes or recipes that we can reuse. The reuse of recipes is already a significant advantage. Additionally, Dataiku is one of the rare platforms that offers end-to-end services where we can carry out a data project from start to finish. For example, to carry out my personal project on predicting a stock index, I did practically everything on Dataiku, from data preparation to putting my model into production.

For example, we have an Excel file that will be incremented every month, a file that comes from outside where the format of the file is exactly the same. Each time we do the transformation, but we did it only on the first file. When the others come in, we apply the old recipes. This allowed us to save an enormous amount of time because we were able to automate everything with Dataiku. As soon as the file comes in, the recipes are automatically applied to these files. We no longer have to intervene at the end of each month and spend twenty to thirty minutes cleaning the files. Additionally, it ensures the validation of our data and consistency between the files.

What I especially prefer about Dataiku are the recipes for data preparation and also the feature we have to create groups of recipes and also to be able to reuse them again.

Once you get the hang of Dataiku, learning the features is also intuitive. It's really a very intuitive platform.

Dataiku is very scalable. It can easily adapt to the expansion of our datasets and it is very powerful. If we have more and more data, Dataiku is very scalable.

Currently, Dataiku is a platform that is almost perfect, and I don't see how to improve it further. I don't have suggestions for potential improvements.

Maybe on the interface in general, the information can easily get lost. If we could summarize the tools bar in a more organized way than what we currently have, that would be helpful.

I have been using Dataiku for one year and a half, and I already obtained my Core Dataiku certification about a month ago.

The licenses are a bit high for companies that are still hesitating to get started with using Dataiku. For my personal projects, I used the thirty-day free trial. Regarding my company, I did not have access to this pricing information.

Given our needs, the best tool, despite the licenses and the cost of the license, Dataiku turns out to be by far the favorite tool compared to the others.

Dataiku is really a very intuitive platform. It allows you to carry out data projects from end to end. We also have the opportunity to reuse templates, models, and recipes. That's one of the big advantages of using Dataiku.

In the context of my personal projects, I developed a pleasure in using Dataiku, which is not the case for other tools. Because the platform is intuitive, I can easily guide myself through it.

I find the documentation on Dataiku very informative and also very instructive.

I would tell others to go for it if Dataiku truly meets their needs. It's the best offer on the market with good documentation. My overall rating for Dataiku is nine out of ten.

My usual use cases for Dataiku are mostly data science use cases, or model creation and training.

The best features of Dataiku that I think are most important for me include the ETL, the Python, processing, database, and API services.

These features are so valuable to me because all these features are available in one place, which helps me create a solution quickly rather than having a couple of technologies together and integrating them.

I have used Dataiku's AutoML tools, and they definitely shorten the process; they have a lot of built-in intelligence which we can use, and they help to automate a lot of things.

Dataiku enhances collaboration within my team because it is one single tool; all the projects are in one place. People can share each other's workbooks and reusable codes.

I think Dataiku could be improved or enhanced in future releases with more 'talk to my data' capabilities, maybe more NLP features, and maybe a platform to build agents.

These improvements would benefit me and my processes because they will help us to continue using Dataiku as one platform; right now we are exploring other platforms for the features which are missing, and if they are available within the same platform, I think it will increase the usage of Dataiku further.

I think the pricing and licensing of Dataiku is a bit expensive; it could be improved further, and I think they should have a different kind of licensing model as well.

I have been working with Dataiku for four years.

As for stability and reliability, so far so good; after the installation, I really had no problems. Dataiku works very well.

Dataiku is quite scalable, as long as I can pay for more licenses, there is no technical limitation.

I would rate the technical support from Dataiku around seven.

I would like them to improve something about the support because it was just our experience that we were complaining about a few things which were not working, and it took a very long time for them to acknowledge that it was not working. It was back and forth, trying this and trying that, and so on. They should not take the complaints so lightly.

Before working with Dataiku, I did not use a different solution for the same use cases; we were doing many things ad hoc, writing notebooks, putting it on, and deploying it to the server ourselves and so on.

The challenges I faced were a lot of technical challenges; there is generally a guide on how to set up and install and configure, and we were following those guides, but we were not able to do it ourselves. When we purchased the licenses, we believed that we could do it ourselves and that it would be a straightforward installation, but then we had to extend the contract with Dataiku to get their experts in order to set up everything. The challenges were mostly that things were not running, there were a lot of technical exceptions, and there was a lot of correlation between the server's version and Dataiku's version. They also had to fix a couple of things on their side because things were not working.

I participated in the initial setup and deployment of Dataiku, and the process was somewhere in the middle; it was not as straightforward as we thought it would be. We had team members who had worked on it before, but we still had to get experts from Dataiku in order to help us, and it took a bit longer than expected.

I consider the return on investment with Dataiku valuable because for us, it is one single platform where all our data scientists come together and work on any model building, so it is collaboration, plus having everything in one place, organized, having proper project management, and then built-in capabilities which help to facilitate model building.

Other than time efficiency and everything being in one place, the other return on investment I observed with Dataiku is that these are the major aspects; if I have to think about taking out Dataiku and putting my own tools and practices in place, it will be a lot of work to build everything together. These for me are the main things.

Before choosing Dataiku, I evaluated other options, specifically Databricks, Dataiku, and our vanilla solutions with Azure and AWS.

The reason I chose Dataiku is that we had team members who had used Dataiku before, which gave us more confidence that we would be successful.

Within the last twelve months, I'm mostly working with Snowflake and Dataiku.

I'm a customer of this solution. I don't know where I purchased Dataiku from.

Dataiku's data source integration flexibility has not benefited my data projects much because we are using our own tools for that.

The valuable insights I have derived from using Dataiku's machine learning capabilities include the fact that we are building our own proprietary model, and Dataiku's machine learning capabilities help us to build those models and create insights. It's very proprietary and we have not used any out-of-the-box insights that are available, but the whole Dataiku application helps to speed up the process.

Dataiku's governance and security controls have helped to maintain data privacy. All those features we are using, and again, this is all part of the applications.

Before Dataiku's implementation, all of this used to be thought of separately and implemented using some tools or technologies, but with Dataiku, all of this comes together in one platform. Once my data is in Dataiku, I know that I can put in security, I can integrate it with my IAM and everything, so I don't have to think of all these features independently. Dataiku provides everything in one platform.

I rate this solution an eight overall.