StarZero All shows Try the podcast clipper →

Apr 13, 2016 · 22m · mad

The Solitude of the Data Team Manager // Florian Douetteau, Dataiku (Hosted by FirstMark)

Florian Douetteau · 17m spoken Matt Turck · 1m 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 presentation and fireside chat at Data Driven NYC, Dataiku Founder & CEO Florian Douetteau explores the operational, technical, and human challenges faced by data team managers. Through a relatable narrative, he outlines practical strategies for unifying data personas, establishing robust architecture, and successfully deploying predictive models into production.

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 7% of the talking time here. How this is scored →

Matt as informed peer 1.0 Guest teaching 1.8 Guest disagreement 0.6 Matt pushing back 0.8
05100:0010:0020:000:14–4:32 · Matt as informed peer 0/10 Florian's Technical Background and the Shift to Visual IDEs This is a solo presentation segment by Florian with no host participation. Host scores are set to zero as required. Florian outlines his technical background and introduces the hypothetical data manager Hal.4:32–8:12 · Matt as informed peer 0/10 Predictive Models in Production, Team Roles, and Mindset Coexistence This is a continuation of the guest's monologue presentation. Host scores are zero due to zero host involvement. Florian explains the need to bridge the gap between clickers and coders.8:12–10:41 · Matt as informed peer 0/10 The Data Disconnect and Data Acquisition Animal Strategies Monologue segment focusing on data acquisition strategies. Host scores remain zero. Florian uses playful animal analogies, noting the devious nature of the Fox strategy.10:41–14:01 · Matt as informed peer 0/10 Product Disconnect, Automated Decision-Making, and API Culture Final segment of the solo presentation. Host scores are zero because the host does not speak. Florian discusses building product features and establishing an API culture.14:01–22:59 · Matt as informed peer 5/10 Wrapping Up Hal's Journey and Dataiku's Mission Matt Turck steps in to host the Q&A session alongside audience members. Matt uses his venture capital background to question Florian on seat monetization, technical barriers, and internal competition, while Florian candidly names SAS as a legacy competitor he intends to displace.0:14–4:32 · Guest teaching 1/10 Florian's Technical Background and the Shift to Visual IDEs This is a solo presentation segment by Florian with no host participation. Host scores are set to zero as required. Florian outlines his technical background and introduces the hypothetical data manager Hal.4:32–8:12 · Guest teaching 1/10 Predictive Models in Production, Team Roles, and Mindset Coexistence This is a continuation of the guest's monologue presentation. Host scores are zero due to zero host involvement. Florian explains the need to bridge the gap between clickers and coders.8:12–10:41 · Guest teaching 2/10 The Data Disconnect and Data Acquisition Animal Strategies Monologue segment focusing on data acquisition strategies. Host scores remain zero. Florian uses playful animal analogies, noting the devious nature of the Fox strategy.10:41–14:01 · Guest teaching 2/10 Product Disconnect, Automated Decision-Making, and API Culture Final segment of the solo presentation. Host scores are zero because the host does not speak. Florian discusses building product features and establishing an API culture.14:01–22:59 · Guest teaching 3/10 Wrapping Up Hal's Journey and Dataiku's Mission Matt Turck steps in to host the Q&A session alongside audience members. Matt uses his venture capital background to question Florian on seat monetization, technical barriers, and internal competition, while Florian candidly names SAS as a legacy competitor he intends to displace.0:14–4:32 · Guest disagreement 0/10 Florian's Technical Background and the Shift to Visual IDEs This is a solo presentation segment by Florian with no host participation. Host scores are set to zero as required. Florian outlines his technical background and introduces the hypothetical data manager Hal.4:32–8:12 · Guest disagreement 0/10 Predictive Models in Production, Team Roles, and Mindset Coexistence This is a continuation of the guest's monologue presentation. Host scores are zero due to zero host involvement. Florian explains the need to bridge the gap between clickers and coders.8:12–10:41 · Guest disagreement 1/10 The Data Disconnect and Data Acquisition Animal Strategies Monologue segment focusing on data acquisition strategies. Host scores remain zero. Florian uses playful animal analogies, noting the devious nature of the Fox strategy.10:41–14:01 · Guest disagreement 0/10 Product Disconnect, Automated Decision-Making, and API Culture Final segment of the solo presentation. Host scores are zero because the host does not speak. Florian discusses building product features and establishing an API culture.14:01–22:59 · Guest disagreement 2/10 Wrapping Up Hal's Journey and Dataiku's Mission Matt Turck steps in to host the Q&A session alongside audience members. Matt uses his venture capital background to question Florian on seat monetization, technical barriers, and internal competition, while Florian candidly names SAS as a legacy competitor he intends to displace.0:14–4:32 · Matt pushing back 0/10 Florian's Technical Background and the Shift to Visual IDEs This is a solo presentation segment by Florian with no host participation. Host scores are set to zero as required. Florian outlines his technical background and introduces the hypothetical data manager Hal.4:32–8:12 · Matt pushing back 0/10 Predictive Models in Production, Team Roles, and Mindset Coexistence This is a continuation of the guest's monologue presentation. Host scores are zero due to zero host involvement. Florian explains the need to bridge the gap between clickers and coders.8:12–10:41 · Matt pushing back 0/10 The Data Disconnect and Data Acquisition Animal Strategies Monologue segment focusing on data acquisition strategies. Host scores remain zero. Florian uses playful animal analogies, noting the devious nature of the Fox strategy.10:41–14:01 · Matt pushing back 0/10 Product Disconnect, Automated Decision-Making, and API Culture Final segment of the solo presentation. Host scores are zero because the host does not speak. Florian discusses building product features and establishing an API culture.14:01–22:59 · Matt pushing back 4/10 Wrapping Up Hal's Journey and Dataiku's Mission Matt Turck steps in to host the Q&A session alongside audience members. Matt uses his venture capital background to question Florian on seat monetization, technical barriers, and internal competition, while Florian candidly names SAS as a legacy competitor he intends to displace.

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

0:00 · Matt 0% · guest 100%0:00 · Matt 0% · guest 100%3:00 · Matt 0% · guest 100%3:00 · Matt 0% · guest 100%6:00 · Matt 0% · guest 100%6:00 · Matt 0% · guest 100%9:00 · Matt 0% · guest 100%9:00 · Matt 0% · guest 100%12:00 · Matt 0% · guest 100%12:00 · Matt 0% · guest 100%15:00 · Matt 31.7% · guest 68.3%15:00 · Matt 31.7% · guest 68.3%18:00 · Matt 24.7% · guest 75.3%18:00 · Matt 24.7% · guest 75.3%21:00 · Matt 1.9% · guest 98.1%21:00 · Matt 1.9% · guest 98.1%
Sharpest disagreement ▶ 18:07 Directly targeting legacy competitor SAS

Florian bluntly names SAS Institute as his primary target, explicitly stating he wants to replace them because he doesn't like SAS.

 Hardest push from Matt ▶ 19:43 Questioning market size and seat density

Matt Turck frames a question from a VC perspective, challenging the economic ceiling of selling exclusively to data scientists versus broader data analyst roles.

 Biggest teaching moment ▶ 21:12 Explaining the operational maintenance bottleneck

Florian educates the room on why data science projects stall in year two due to model maintenance, explaining why dedicated data ops roles are required.

 Matt holds his own ▶ 19:43 Demonstrating VC domain knowledge

Matt Turck demonstrates domain expertise in software monetization models by probing into seat counts and buyer personas within enterprise accounts.
the scores for every segment, with the reasoning behind each
ChapterTopicMatt as informed peerGuest teachingGuest disagreementMatt pushing backWhy
Florian's Technical Background and the Shift to Visual IDEs 0100 This is a solo presentation segment by Florian with no host participation. Host scores are set to zero as required. Florian outlines his technical background and introduces the hypothetical data manager Hal.
Predictive Models in Production, Team Roles, and Mindset Coexistence 0100 This is a continuation of the guest's monologue presentation. Host scores are zero due to zero host involvement. Florian explains the need to bridge the gap between clickers and coders.
The Data Disconnect and Data Acquisition Animal Strategies 0210 Monologue segment focusing on data acquisition strategies. Host scores remain zero. Florian uses playful animal analogies, noting the devious nature of the Fox strategy.
Product Disconnect, Automated Decision-Making, and API Culture 0200 Final segment of the solo presentation. Host scores are zero because the host does not speak. Florian discusses building product features and establishing an API culture.
Wrapping Up Hal's Journey and Dataiku's Mission 5324 Matt Turck steps in to host the Q&A session alongside audience members. Matt uses his venture capital background to question Florian on seat monetization, technical barriers, and internal competition, while Florian candidly names SAS as a legacy competitor he intends to displace.

Statements from this episode (9)

Insight
Douetteau: Data leaders must choose between polyglot stacks or dictating standards
“And so he, possibly both are fine, I think, but he really need to choose, because you cannot really be a polyglot dictator.”
Florian Douetteau Apr 13, 2016 ▶ 4:19
Prediction Not checkable as stated
Douetteau: Winning companies will excel at putting predictive models into production
“So it might be that in the following decade, about, where everything will be about having intelligent connected devices, a company that will win will be companies that will be very good at putting predictive models in production, putting intelligence in produc…”
Florian Douetteau Apr 13, 2016 ▶ 4:54
Insight
Douetteau: Effective data teams must bridge visual interfaces and programming
“In my perspective, well, whatever happens today, you must be able to make the two mindsets coexist, meaning when you work with computers today, you have people which are more into the coding kind of mode, meaning a textual interaction with computers, and peopl…”
Florian Douetteau Apr 13, 2016 ▶ 7:36
Insight
Douetteau: The main reason big data projects fail is lacking data
“I didn't really make the stats, but I think the main reason for a big data project to fail is not having the data.”
Florian Douetteau Apr 13, 2016 ▶ 8:20
Insight
Douetteau: Algorithmic optimization accounts for only 20% of a data project
“When you do have the data, and you do have the real business problem you want to solve, it's like you already did, like, half of the job, and so maybe you spend, like, more, 30 more percent actually working on the domain-specific or problem-specific features y…”
Florian Douetteau Apr 13, 2016 ▶ 8:41
Insight
Douetteau: Data teams must include or work closely with product managers
“So, in my perspective, in order to do that, a data team must be kind of close to the product team, meaning he really needs to have either someone with a product management mindset, either in the team or very, very close to the team, because any significant dat…”
Florian Douetteau Apr 13, 2016 ▶ 12:12
Opinion
Douetteau: Dataiku aims to replace SAS Institute because "I don't like SAS"
“I consider my main competitor to be the SAS, SAS Institute, meaning, ah, actually that's the company I want to replace, because I don't like SAS.”
Florian Douetteau Apr 13, 2016 ▶ 18:09
Disclosure
Douetteau: Dataiku has approximately 30 active user seats at AXA
“AXA, AXA is our customer for about, ah, two years, I think. And so, they are using our software internally in their, ah, data initiative. So yeah, we have about 30 seats of, at AXA, 30 users at AXA, I think.”
Florian Douetteau Apr 13, 2016 ▶ 19:24
Prediction Not checkable as stated
Douetteau: Future data ops roles will bridge system scalability and ML metrics
“I think a challenge would be in the future to have, like specialized people for operations on data that will need to understand well scalability issues from a system administrator perspective, but also what F-one score is, because they will need to be, ah, sma…”
Florian Douetteau Apr 13, 2016 ▶ 22:26
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