nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Dataiku

Peter_Polito

I am working with a prospective customer currently using Dataiku. Dataiku seems to be more of an enterprise product but need to be able to speak directly to how JMP is different than Dataiku. Anyone with experience with this product and can speak to how we can differentiate ourselves? Thanks.

Accepted answers

All comments

ian_jmp

Perhaps you have seen the recent thread about the NXP Hackathon. FWIW, 'DataIku' is the tool that NXP seems to have standardised on. As a member of the judging panel, I saw several demos. Others involved with this project may have their own commentary, but my summary would be that, although JMP could co-exist (synergistically . . . ) with DataIku, it could never replace it.

Daniel_Valente

We just ran into them at another customer. My impression is pretty similar here. Different tools for different tasks. Limited analytics with the data science bundle, which is an extra cost, and allows machine learning and data mining through Python. Seems like everything would need to be configured and developed for a specific use-case, vs. the general problem solving tool that is JMP. In this customer interaction, it was actually suggested that Dataiku could be used for data cleanup and warehousing tasks ahead of JMP analysis as well (connecting to live data sources) so this certainly isn’t and either/or conversation.

Daniel_Valente

@Emmanuel_Romeu may be able to provide some additional information on the customer interaction that I was speaking about above. It was my take-away from the call that there is some overlap, but the use-cases organizationally would be different for Dataiku and JMP. JMP Live has a bigger overlap with Dataiku, especially with plans for data refresh capabilities server-side in JMP 17 and beyond.

Emmanuel_Romeu

We started the conversation about SEEQ vs. JMP and they also talked about Dataiku. In conclusion, it seems that they need SEEQ or Dataiku for real-time monitoring capabilities. JMP offers better model development capability compared to SEEQ but in an easier way compared to Dataiku. Dataiku has good AI/ML capabilities but requires higher skills.

ChulHee_JMP

Hi Peter,

I'm wondering if you have any updtates about the prospective customer.

Currently, there Dataiku has been spreading in Korea as DX(Digital Transformation) or AI initiative tool.

That's why I'm curious about it.

Thanks!

russ_wolfinger

Wanted to mention that the XGBoost and Torch add-ins level up JMP Pro for AI/ML initiatives. We have customers already using Torch to build and deploy models in Python.

Also a personnel note: For the past couple years Kelci Miclaus has been Dataiku's Global Head for AI/ML Life Sciences. She was formerly a star intern and then coder and manager on our JMP LIfe Sciences R&D team. So they are certainly well aware of us and are positioning accordingly. The Torch add-in has just been released this year and they likely now view JMP Pro as direct competition to what they do with Python wrappers.

ChulHee_JMP

Thanks for your comments, @russ_wolfinger !

I have observed that many Korean companies recently adopt and implement DT (Digital Transformation) initiative including AI/ML techniques. I think that a good example of "truly transformational" DT would be Tesla's attempts related with IoT and EV car.

Meanwhile, because it takes a long time to develop a data scientist like you, companies are trying to train all of their scientists and engineers to become citizen data scientists in a short period of time. At the same time, they look for tools that enable data analysis with no-code or low-code. Learning R and/or Python and applying them into their works well is too much burden to scientists and engineers in general. So, the tool should be able to perform artificial intelligence and machine learning more easily. In a recent interview with a Dataiku user from a large Korean company, it seems that Dtaiku has largely satisfied their needs in this respect. It is characterized by allowing the workflow of the entire AI/ML project process - from data collection and cleaning to modeling - to be seen visually at a glance, as if it were based on JMP's Reliability Block Diagram and various functions of File and Tables. It appears to flexibly integrate DB access and data organization, visualization of Graph Builder, and even analysis and modeling using Torch Add-in. Additionally, it seems that Dataiku's collaboration function is similar to correspond to the integration of JMP's Project function and JMP Live.

nick_shelton

@saradoudt We seem to be encountering more and more conversations with customers/prospects regarding Dataiku's capabilities versus JMP. Would be valuable to have a competitor sales sheet (similar to what's available for Minitab) for Dataiku. I don't think the details that have been captured for Dataiku on the competitor page (JMP Competitors - Dataiku) are sufficient to prepare teams for these discussions.

Peter_Polito

I have since changed teams and no longer work with that customer. Regardless, I am running into Dataiku frequently in the semiconductor space. A few weeks ago I attended SemiconWest and of the talks I attended more mentioned Dataiku than did not. It is a competitor we need to intimately understand and develop talking points to both differentiate JMP from Dataiku and identify complementarity. Dataiku is rapidly expanding and without the angle of complementarity, I think we will have a more and more difficult time making inroads where it is established, which is further punctuated by the enterprise nature of that tool.