StarZero All shows Try the podcast clipper →

Apr 27, 2021 · 27m · mad

Fireside Chat: Florian Douetteau (Founder & CEO, Dataiku) with Matt Turck (Partner, FirstMark)

Florian Douetteau · 17m spoken Matt Turck · 6m spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

In this fireside chat hosted by FirstMark's Matt Turck, Dataiku Founder and CEO Florian Douetteau discusses Dataiku's journey from a French startup to an enterprise AI unicorn, detailing platform design principles, cloud partnerships, and the democratization of machine learning.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 26.7% of the talking time here. How this is scored →

Matt as informed peer 3.0 Guest teaching 3.2 Guest disagreement 0.5 Matt pushing back 0.0
05100:0010:0020:004:07–8:48 · Matt as informed peer 2/10 Defining Enterprise AI and Key Use Cases Matt opens with a broad introductory prompt asking about enterprise AI definition and use cases. Florian gently reframes the notion of enterprise AI, explaining that it is not about magical products but a 20-25 year journey of optimizing mundane business processes. Florian also corrects the premise comparing enterprises to tech giants, noting enterprises should not copy FAANG by spending hundreds of millions on custom platforms.8:48–14:04 · Matt as informed peer 5/10 Platform Design Principles and Democratization Matt demonstrates substantial insider domain knowledge as an early investor, stepping in to articulate the platform's multi-persona collaboration layer and system of record functionality. Florian builds upon this by detailing how Dataiku bridges visual analysts and Python coders to eliminate operational silos. The exchange is highly collaborative with zero pushback or friction.14:04–16:13 · Matt as informed peer 4/10 Competitive Landscape and Company Identity Matt brings up key market competitors from audience Q&A like Databricks and DataRobot to explore Dataiku's positioning. Florian educates the audience on the distinction, explaining that Databricks focuses on infrastructure and DataRobot on AutoML, while Dataiku focuses on human creativity across the lifecycle. Matt also playfully prompts Florian on the origin of the company name.16:13–19:51 · Matt as informed peer 3/10 Strategic Partnership with Snowflake and Dataiku Online Matt highlights the timely partner announcement with Snowflake and prompts discussion around Dataiku Online. Florian details the technical integration with Snowpark and how cloud accessibility lowers the friction of running ML pipelines. The conversation remains entirely supportive and promotional.19:51–23:44 · Matt as informed peer 2/10 Operational Insights and Dataiku 9 Capabilities Matt references Florian's recent article to ask about operational lessons learned over eight years. Florian explains technical features in Dataiku 9, such as machine learning assertions, designed to bridge statistical models with domain-specific business knowledge.23:44–27:01 · Matt as informed peer 2/10 Q&A on Data Privacy and Global Expansion Matt moderates final Q&A questions regarding data privacy and US market expansion. Florian answers pragmatically about differential privacy and jokingly notes that moving to the US was necessary because that is where the appetite for enterprise AI resides.4:07–8:48 · Guest teaching 4/10 Defining Enterprise AI and Key Use Cases Matt opens with a broad introductory prompt asking about enterprise AI definition and use cases. Florian gently reframes the notion of enterprise AI, explaining that it is not about magical products but a 20-25 year journey of optimizing mundane business processes. Florian also corrects the premise comparing enterprises to tech giants, noting enterprises should not copy FAANG by spending hundreds of millions on custom platforms.8:48–14:04 · Guest teaching 3/10 Platform Design Principles and Democratization Matt demonstrates substantial insider domain knowledge as an early investor, stepping in to articulate the platform's multi-persona collaboration layer and system of record functionality. Florian builds upon this by detailing how Dataiku bridges visual analysts and Python coders to eliminate operational silos. The exchange is highly collaborative with zero pushback or friction.14:04–16:13 · Guest teaching 4/10 Competitive Landscape and Company Identity Matt brings up key market competitors from audience Q&A like Databricks and DataRobot to explore Dataiku's positioning. Florian educates the audience on the distinction, explaining that Databricks focuses on infrastructure and DataRobot on AutoML, while Dataiku focuses on human creativity across the lifecycle. Matt also playfully prompts Florian on the origin of the company name.16:13–19:51 · Guest teaching 3/10 Strategic Partnership with Snowflake and Dataiku Online Matt highlights the timely partner announcement with Snowflake and prompts discussion around Dataiku Online. Florian details the technical integration with Snowpark and how cloud accessibility lowers the friction of running ML pipelines. The conversation remains entirely supportive and promotional.19:51–23:44 · Guest teaching 3/10 Operational Insights and Dataiku 9 Capabilities Matt references Florian's recent article to ask about operational lessons learned over eight years. Florian explains technical features in Dataiku 9, such as machine learning assertions, designed to bridge statistical models with domain-specific business knowledge.23:44–27:01 · Guest teaching 2/10 Q&A on Data Privacy and Global Expansion Matt moderates final Q&A questions regarding data privacy and US market expansion. Florian answers pragmatically about differential privacy and jokingly notes that moving to the US was necessary because that is where the appetite for enterprise AI resides.4:07–8:48 · Guest disagreement 1/10 Defining Enterprise AI and Key Use Cases Matt opens with a broad introductory prompt asking about enterprise AI definition and use cases. Florian gently reframes the notion of enterprise AI, explaining that it is not about magical products but a 20-25 year journey of optimizing mundane business processes. Florian also corrects the premise comparing enterprises to tech giants, noting enterprises should not copy FAANG by spending hundreds of millions on custom platforms.8:48–14:04 · Guest disagreement 0/10 Platform Design Principles and Democratization Matt demonstrates substantial insider domain knowledge as an early investor, stepping in to articulate the platform's multi-persona collaboration layer and system of record functionality. Florian builds upon this by detailing how Dataiku bridges visual analysts and Python coders to eliminate operational silos. The exchange is highly collaborative with zero pushback or friction.14:04–16:13 · Guest disagreement 1/10 Competitive Landscape and Company Identity Matt brings up key market competitors from audience Q&A like Databricks and DataRobot to explore Dataiku's positioning. Florian educates the audience on the distinction, explaining that Databricks focuses on infrastructure and DataRobot on AutoML, while Dataiku focuses on human creativity across the lifecycle. Matt also playfully prompts Florian on the origin of the company name.16:13–19:51 · Guest disagreement 0/10 Strategic Partnership with Snowflake and Dataiku Online Matt highlights the timely partner announcement with Snowflake and prompts discussion around Dataiku Online. Florian details the technical integration with Snowpark and how cloud accessibility lowers the friction of running ML pipelines. The conversation remains entirely supportive and promotional.19:51–23:44 · Guest disagreement 0/10 Operational Insights and Dataiku 9 Capabilities Matt references Florian's recent article to ask about operational lessons learned over eight years. Florian explains technical features in Dataiku 9, such as machine learning assertions, designed to bridge statistical models with domain-specific business knowledge.23:44–27:01 · Guest disagreement 1/10 Q&A on Data Privacy and Global Expansion Matt moderates final Q&A questions regarding data privacy and US market expansion. Florian answers pragmatically about differential privacy and jokingly notes that moving to the US was necessary because that is where the appetite for enterprise AI resides.4:07–8:48 · Matt pushing back 0/10 Defining Enterprise AI and Key Use Cases Matt opens with a broad introductory prompt asking about enterprise AI definition and use cases. Florian gently reframes the notion of enterprise AI, explaining that it is not about magical products but a 20-25 year journey of optimizing mundane business processes. Florian also corrects the premise comparing enterprises to tech giants, noting enterprises should not copy FAANG by spending hundreds of millions on custom platforms.8:48–14:04 · Matt pushing back 0/10 Platform Design Principles and Democratization Matt demonstrates substantial insider domain knowledge as an early investor, stepping in to articulate the platform's multi-persona collaboration layer and system of record functionality. Florian builds upon this by detailing how Dataiku bridges visual analysts and Python coders to eliminate operational silos. The exchange is highly collaborative with zero pushback or friction.14:04–16:13 · Matt pushing back 0/10 Competitive Landscape and Company Identity Matt brings up key market competitors from audience Q&A like Databricks and DataRobot to explore Dataiku's positioning. Florian educates the audience on the distinction, explaining that Databricks focuses on infrastructure and DataRobot on AutoML, while Dataiku focuses on human creativity across the lifecycle. Matt also playfully prompts Florian on the origin of the company name.16:13–19:51 · Matt pushing back 0/10 Strategic Partnership with Snowflake and Dataiku Online Matt highlights the timely partner announcement with Snowflake and prompts discussion around Dataiku Online. Florian details the technical integration with Snowpark and how cloud accessibility lowers the friction of running ML pipelines. The conversation remains entirely supportive and promotional.19:51–23:44 · Matt pushing back 0/10 Operational Insights and Dataiku 9 Capabilities Matt references Florian's recent article to ask about operational lessons learned over eight years. Florian explains technical features in Dataiku 9, such as machine learning assertions, designed to bridge statistical models with domain-specific business knowledge.23:44–27:01 · Matt pushing back 0/10 Q&A on Data Privacy and Global Expansion Matt moderates final Q&A questions regarding data privacy and US market expansion. Florian answers pragmatically about differential privacy and jokingly notes that moving to the US was necessary because that is where the appetite for enterprise AI resides.

speaking balance: gold is Matt, purple is the guest (3 minute bins)

0:00 · Matt 64.8% · guest 35.2%0:00 · Matt 64.8% · guest 35.2%3:00 · Matt 18.3% · guest 81.7%3:00 · Matt 18.3% · guest 81.7%6:00 · Matt 18.1% · guest 81.9%6:00 · Matt 18.1% · guest 81.9%9:00 · Matt 17.7% · guest 82.3%9:00 · Matt 17.7% · guest 82.3%12:00 · Matt 35.5% · guest 64.5%12:00 · Matt 35.5% · guest 64.5%15:00 · Matt 11.2% · guest 88.8%15:00 · Matt 11.2% · guest 88.8%18:00 · Matt 38.4% · guest 61.6%18:00 · Matt 38.4% · guest 61.6%21:00 · Matt 9.8% · guest 90.2%21:00 · Matt 9.8% · guest 90.2%24:00 · Matt 20.5% · guest 79.5%24:00 · Matt 20.5% · guest 79.5%27:00 · Matt 82.7% · guest 17.3%27:00 · Matt 82.7% · guest 17.3%
Sharpest disagreement ▶ 6:39 Reframing the FAANG comparative framing

Florian gently rejects the premise that enterprises need to emulate Google or Facebook by building custom platforms, arguing that enterprises have better things to do than spend hundreds of millions on custom platform engineering.

 Hardest push from Matt ▶ 14:42 Interrupting to query company naming

In an otherwise completely non-confrontational conversation, Matt playfully interrupts Florian to push him to explain what the 'haiku' in Dataiku actually stands for.

 Biggest teaching moment ▶ 14:51 Clear positioning against major market competitors

Florian succinctly educates the audience on market categorization, contrasting Databricks (infrastructure/plumbing) and DataRobot (AutoML point solution) against Dataiku's human-creativity end-to-end approach.

 Matt holds his own ▶ 11:43 Demonstrating board-level understanding of platform value

Matt steps in to explain the exact mechanics of Dataiku's cross-persona collaboration features and multi-location system of record functionality, demonstrating deep domain expertise as an early investor.
the scores for every segment, with the reasoning behind each
ChapterTopicMatt as informed peerGuest teachingGuest disagreementMatt pushing backWhy
Defining Enterprise AI and Key Use Cases 2410 Matt opens with a broad introductory prompt asking about enterprise AI definition and use cases. Florian gently reframes the notion of enterprise AI, explaining that it is not about magical products but a 20-25 year journey of optimizing mundane business processes. Florian also corrects the premise comparing enterprises to tech giants, noting enterprises should not copy FAANG by spending hundreds of millions on custom platforms.
Platform Design Principles and Democratization 5300 Matt demonstrates substantial insider domain knowledge as an early investor, stepping in to articulate the platform's multi-persona collaboration layer and system of record functionality. Florian builds upon this by detailing how Dataiku bridges visual analysts and Python coders to eliminate operational silos. The exchange is highly collaborative with zero pushback or friction.
Competitive Landscape and Company Identity 4410 Matt brings up key market competitors from audience Q&A like Databricks and DataRobot to explore Dataiku's positioning. Florian educates the audience on the distinction, explaining that Databricks focuses on infrastructure and DataRobot on AutoML, while Dataiku focuses on human creativity across the lifecycle. Matt also playfully prompts Florian on the origin of the company name.
Strategic Partnership with Snowflake and Dataiku Online 3300 Matt highlights the timely partner announcement with Snowflake and prompts discussion around Dataiku Online. Florian details the technical integration with Snowpark and how cloud accessibility lowers the friction of running ML pipelines. The conversation remains entirely supportive and promotional.
Operational Insights and Dataiku 9 Capabilities 2300 Matt references Florian's recent article to ask about operational lessons learned over eight years. Florian explains technical features in Dataiku 9, such as machine learning assertions, designed to bridge statistical models with domain-specific business knowledge.
Q&A on Data Privacy and Global Expansion 2210 Matt moderates final Q&A questions regarding data privacy and US market expansion. Florian answers pragmatically about differential privacy and jokingly notes that moving to the US was necessary because that is where the appetite for enterprise AI resides.

Statements from this episode (11)

Disclosure
Douetteau: Dataiku has 650 employees and is doubling headcount annually
“We are six, 50 employees and doubling kind of on a year to year basis.”
Florian Douetteau Apr 27, 2021 ▶ 3:16
Prediction Not checkable as stated
Douetteau: Enterprise AI adoption is probably a 20 to 25 year journey
“It's probably a 20, 25 years journey, and we are like one third into it.”
Florian Douetteau Apr 27, 2021 ▶ 5:11
Prediction Not checkable as stated
Douetteau: Most companies won't spend $200M-$500M building custom AI platforms
“Meaning they won't spend two hundred millions or five hundred millions in order to build from scratch a data and AI platform just because it's not realistic.”
Florian Douetteau Apr 27, 2021 ▶ 8:20
Prediction Not checkable as stated
Douetteau: Enterprises won't use five to eight data tools in 10 years
“There is no way for the enterprise in the long term, if you move forward and put yourself from a, if you go forward by 10 years, that you have like five, six, seven, eight different tools that you use to manage the data life cycle. It will make no sense.”
Florian Douetteau Apr 27, 2021 ▶ 9:19
Insight
Douetteau: Global supply of data scientists cannot meet demand for data problems
“There are not enough data scientists in the world to solve all the data problems we have.”
Florian Douetteau Apr 27, 2021 ▶ 11:04
Opinion
Douetteau: Databricks is mostly focused on data infrastructure and placement
“Companies such as the Databricks of the world are mostly about infrastructure and actually helping you put data In a given location.”
Florian Douetteau Apr 27, 2021 ▶ 15:12
Opinion
Douetteau: DataRobot focuses primarily on AutoML and ML automation
“Company like DataRobot are mostly about automating and especially the ML part of the process and meaning the AutoML part of the process.”
Florian Douetteau Apr 27, 2021 ▶ 15:21
Assertion Not checkable as stated
Douetteau: ML on gigabytes of data moved from days to clicks
“Meaning a few years ago doing all those EV lifting of being able to do ML on a few gigabytes of data was possibly taking a few days to any seasoned data engineers. Today, you can actually do the same kind of achievement in a few clicks”
Florian Douetteau Apr 27, 2021 ▶ 18:25
Insight
Douetteau: ML model validation rules will not happen without business-oriented platforms
“In order to make it happen, you need indeed to facilitate for business users the ability to just write those rules so that when a data scientist is building a machine learning model, those rules are actually checked. That's where actually the value comes becau…”
Florian Douetteau Apr 27, 2021 ▶ 22:46
Insight
Douetteau: Model auditing and regulation are bigger enterprise bottlenecks than ML performance
“In, in some use cases, it's not machine learning per se, or the performance of machine learning models that is the choke point in order to deliver value. It's the ability to actually meet the, meaning the regulatory, the regulation constraints, meaning literal…”
Florian Douetteau Apr 27, 2021 ▶ 24:34
Opinion
Douetteau: The US holds the greatest concentration of data and AI ambition
“Literally speaking, where are the treasure troves of data and AI? Where are the people having the will to do more with data and AI? I must admit, and it will be painful, but indeed it's in the US.”
Florian Douetteau Apr 27, 2021 ▶ 26:21
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 400 conversations transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more