This week, we are excited to release our new fully managed offering of Dataiku: Dataiku Online. To dig a bit deeper into the new release, I sat down with Louis-Philippe Kronek, the General Manager of Dataiku Online, to discuss why and how his team developed this new offering over the past few months.
Why develop Dataiku Online, a fully managed version of Dataiku?
Honestly, we were just listening to our customers. There was a strong demand from customers for a fully managed offering of Dataiku. This was especially true amongst small and mid-sized businesses (as well as startups) who predominantly use managed services and cloud data. These teams and organizations don’t have dedicated IT support, so having Dataiku take care of all the infrastructure and upgrades makes a lot of sense, plus it is just easier to get started and create value right away.
We chose to adopt Dataiku in SaaS mode to free ourselves from routine support and maintenance tasks. This means we can focus on very quickly implementing our clients' data projects without constraints on data volumes or of modeling complexity.”
Francois Fau, Lead Data Scientist at June Partners
Can you tell me a bit about the first customers that have started using Dataiku Online and how they differ from other Dataiku customers?
Dataiku Online was developed for small and mid-sized companies that don't rely as heavily on on-premise technologies or custom clouds like larger enterprise companies do. At the beginning of the year, we engaged with a number of companies who expressed interest in this type of offering and the results have been great. What they have in common is that many are at the beginning of their journey with data analytics and want to develop machine learning (ML) and AI-powered services. They’ve solved the initial problems of data ingestion and have broken down silos between their data, and now they’re looking to actually leverage that data and go further in their journey with ML.
In many ways we’re actually talking to companies that are quite similar to the first clients of Dataiku (Louis-Philippe has been working with Dataiku for nearly seven years now — he would know). Their tech and data stacks have changed though — while we used to talk to people with data lakes, we’re now seeing companies that work mainly with SaaS tools and are looking to gather that data together.
Dataiku Online has rapidly enabled us to work collaboratively with a variety of sources of data and will drive models to production much faster than if we had used other tools.”
Eric Buras, Head of Data Science and ML, Bidops
I assume there are many similarities as well between users of Dataiku Online and installed options of Dataiku?
Yes, of course! Fundamentally they are interested in Dataiku for the same reasons, and having the software available as a service is just an enabler to make accessing it easier. They’re looking for an end-to-end analytics and data science platform with all the tools necessary to create data-driven products. They’re also looking to enable people with different skill sets. Working with data is often not a full-time job in the hands of a single person in these companies.
Data usage is split between different profiles with different backgrounds who can clean and combine data, build analytics reports and dashboards, and even build ML models. I think many of these new users are surprised at how easy it is for them to collaborate as they work on data projects in Dataiku Online, which is something that our larger customers value highly and take advantage of every day.
The good news is that we offer the same features in Dataiku Online as we have in our installed version. Dataiku Online has a specific configuration and design choices for smaller teams and for managed use. We are actively working with Dataiku Online customers to make sure they have the features they need all while keeping the experience streamlined and easy to use.
When you talk to our customers, what are the biggest improvements that Dataiku Online brings to our customers?
We have just started working with customers over the past few months. What we hear the most is that Dataiku Online has allowed them to get started very quickly. They have a much shorter time to value. It takes them minutes to create an account and a few hours to get up and running and set up connections, so they’re able to deliver their first assets very quickly.
While we thought that the biggest improvement for our clients would be the managed part of the service, we actually found out that for our customers, it’s not even part of the value proposition — it’s a no brainer, a prerequisite. These companies are fully cloud-native companies building out their modern data stack with all managed products — FiveTran, Snowflake, connecting to SaaS business tools like Salesforce or HubSpot, etc. Dataiku Online naturally fits on top of that to fully leverage the data they have collected.
Dataiku Online allows us to focus on analysis, not server administration. The customer service is fast as well. Data insights fuel our growth, and Dataiku Online enables us to develop insights faster than our competitors."
Scott Walker, Managing Partner, Sarissa Partners
Was there anything specific about how we worked to build Dataiku Online?
What made it possible for us to go from dev to launch in under a year is that we started out with a small team that was mostly technical but that also included a business development person from day one (Hugo Le Squeren, our Director of Business Development). That allowed us to work on our go-to-market strategy at the same time as the design and development of the platform. Our goal was to build a new offering, not just a product, and that really drove the whole process.
When we had our first Minimum Viable Product (MVP), we were able to put it in front of real clients really fast and align it with their expectations of a self-service Dataiku. Having a business goal from the start with someone owning it was a real asset.
From the start, we worked with Snowflake to fully integrate Dataiku Online with their partner ecosystem. Can you tell me a bit about why that was important?
We had an opportunity to work with Snowflake to develop an integration through their Partner Connect to access Dataiku in one click on Snowflake. It was a great opportunity to face users that had a cloud mindset, since Snowflake is widely known for making cloud warehousing frictionless and self service. It just seemed obvious that the two offers matched up.
Data integration is key to the overall success of our tech stack and Dataiku was easy to connect to Snowflake, resulting in a seamless data flow between the two. With such a clean connection, we are able to fully leverage the results of our data scientists and easily automate their models.”
Mark Sturino, VP of Data and Analytics, Good Apple
We also recently added additional integrations to Amazon Redshift and Google Big Query. Dataiku has always had a focus on being an agnostic platform and we realize how critical it is for our users to be able to break down silos and access all their data in one place, whether it’s stored in a database or in a SaaS tool. That will continue to be our focus in the coming months: making Dataiku Online more accessible and more open to more companies and users.