CTO Sessions: an interview with Clément Stenac, Dataiku
Wondering what type of CTO is Dataiku's CTO? Read this interview from IDG Connect to find out.
Engineering at Dataiku
A look into how the R&D team works at Dataiku and what makes it special. TLDR: We prioritize quality of product and people's lives over deadlines.
See open positionsWe Are Building the Future of Working With AI
“If you don’t know where you want to go, then it doesn’t matter which path you take.”
— Alice in Wonderland
Dataiku’s story began in Paris in 2013 with a forward-thinking team of four entrepreneurs convinced that in order to succeed in the world’s rapidly evolving ecosystem, organizations — no matter their industry or size — must use data to continuously innovate. Our CEO Florian Douetteau and our CTO Clément Stenac, both co-founders, are experienced developers who believe great products solve complex problems.
We Work Together to Achieve our Common Goal
“I have always believed helping your fellow man is profitable in every sense, personally and bottom-line.”
— Michael Corleone, The Godfather
At Dataiku we all do “product first” — our developers don’t do “code first,” designers don’t do “design first,” and QA don’t do “tests first.” We all strive to achieve the best possible outcome for the product while also working together and fostering healthy teams.
We develop best-in-class products
Our software engineers (split into eight teams with dedicated scopes like ML, data viz, architecture, etc.), software engineers in test, product managers, and UX designers — all based out of Europe with the main hub in Paris — work together to build the best Everyday AI product.
We help our users
Our field engineers, architects, and technical support engineers — with the support of sales and customer success teams — help customers from all around the world, whether it’s with installation, training, or answering day-to-day questions.
We prepare for what’s next
Research scientists from The Lab investigate emerging AI & ML topics, build prototypes, write publications, and help prepare the future of our products. The Incubator team unveils new usages and explores promising product ideas.
Backend and Frontend
Devops
AI & Machine Learning
IDE
Tools
We Value Simple Over Shiny and Efficient Over Trendy
“It is not our abilities that show what we truly are… it is our choices.”
— Albus Dumbledore, Harry Potter & the Chamber of Secrets
Our tech stack (made up of the hottest data and AI tech, including Spark, Kubernetes, deep learning, and more) reflects who we are: a fast-growing company committed to integrating with the latest and greatest — when it makes sense — to provide the best product for our customers. Our backend is mainly written in Java but also includes some Python and bits of Scala and R. Our frontend is based on AngularJS, Angular and also makes vast usage of d3.js for interactive charts.
A Word About Open-source
You can think of our product as a control room for the more than 30 open-source tools with which the platform integrates. Most of our in-house plugins are also open source (Apache License). As a team, we contribute to major projects including:
Dataiku Lab is the author and maintainer of this Python package designed to perform and monitor active learning experiments, leveraging various query sampling methods and metrics.
We actively sponsor this Javascript code editor.
We are part of the first consortium of corporate sponsors who support the development of this flagship machine learning library.
We Believe Diversity of People and Opinions Make Healthy Teams and Great Products
“When we see our uniqueness as a virtue, only then will we find peace.”
– Aramaki, Ghost in the Shell
Dataikers share a common conviction that no universal product, nor healthy teams, can be built without the strength of diverse point of views, social origins, gender, experiences and cultures. We all continuously work on building a diverse and inclusive work environment where all Dataikers feel free to be their true selves (ie: implementing ERG groups, nominating an employee-led DEI committee, and supporting events such as The Grace Hopper Celebration).
We Care About People and Quality More Than Deadlines and Delivery
“All we have to decide is what to do with the time that is given us.”
— Gandalf, the Lord of The Rings
Real Flexibility
Want a vacation? Take it (with or without much notice — don't worry, your team will cope with your absence!). The same goes for remote work and parental leave and working remotely or in the office — we're flexible. If you feel like spending some time volunteering, select a project you cherish and use the Ikig.AI (AI for Good) days you accrue every year.
Responding to Change
We don’t work with artificial deadlines — teams are assigned a scope of features for our releases, and we develop them as quickly as we can. However, we ultimately own the timeline, which means flexibility to delay product feature releases when it's the right thing to do for our customers and product.
Non-Stop Learning
We develop new skills outside of our theoretical area of expertise by sharing our knowledge informally with colleagues, chatting in the kitchen office, answering colleagues’ questions on Slack, or attending improvised whiteboard sessions. We also organize internal workshops and weekly discussions (Tuesday Lunches).
Meet Some of the Team
-
Regis
Senior Software Engineer in Test
-
Agathe
Software Engineer
-
Simona
Senior Research Scientist
We Are Agile in Our Own Way to Maximize Autonomy & Flexibility
“No need to report to him until we have something to report!”
— Nute Gunray, Star Wars: Episode I — The Phantom Menace
This is how we work everyday, and the Hackaiku week (an internal event held every year) is a good representation of this. The only rule is “do whatever you want” — results include rebuilding our Jenkins pipelines, writing a VS Code extension for Dataiku, or detecting and warning users of common errors in ML pipelines.
Pragmatism and flexibility are our driving principles. Rather than rigorously following a pre-established development methodology, we evolved our own based on existing ones, agile principles, and a number of practical issues we have observed in the past. We encourage all teams to adopt this process (but also let them customize it to their needs).
We avoid silos: People can easily express their desire to work on specific projects, and we encourage everyone to explore different technologies and aspects of the product.
We are autonomous: We trust our engineers and expect them to find their way through the code base by themselves as much as possible. This being said, we are always here to help each other as we don’t see the value of “getting stuck.” With this comes a lot of leverage (as well as responsibility) for technical choices.
We are impactful: At Dataiku, no matter your role, you are allowed to suggest, try out new things, and fail. As engineers, we not only have a say in how we do things but also in what we do.
Our View on Hot Topics
What I Wish I Had Known Before Developing a Geospatial Analysis Feature
Our engineers recently worked on implementing a geo join in Dataiku and they learned a ton about geospatial analysis during that process.
Lessons Learned Shadowing the Mythical 10x Programmer
Marine Sobas, junior software engineer, shadowed two senior developers at Dataiku — here's what she learned.
Apply. Interview. Learn: Dataiku’s Engineering Hiring Philosophy
Four main principles guide our engineering hiring — here’s a look at them.
Distributed Hyperparameter Search: How It’s Done in Dataiku
Get to know how our engineering team built Dataiku's new Distributed Hyperparameter Search feature.
Hackaiku: When Dataiku Engineers Turn Into Data Hackers
Every year, the Dataiku engineers participate in Hackaiku week, our annual internal hackathon! See how the event works and uncover this year's top three projects.
Building a Graph Visualization Tool
In this project walk-through, visualize a table as a graph network to get insights about complex connected data — the Game of Thrones characters.
SaaS Rebels With a Cause
Check out this blog highlighting the key takeaways from Dataiku CTO Clément Stenac’s “SaaS Rebels: Why ‘On-Premises’ Is Still Alive and Here to Stay” presentation.
{{ position.location.name }}