DataRobot raises $21M Series A | Hacker News Reader

DataRobot raises $21M Series A

blogs.wsj.com

81 pointsbenhamner8 comments

Jeremy from DataRobot here. Thanks for the support! We're looking forward to releasing the product and showing off all of our hard work!

Simone: Thanks for the congrats and all your help during and after techstars.

NnamdiJr: I did mention Xavier to the wsj--not sure why they didn't include him in the article. Xavier has been working with us full time for close to a year now. Him and I have been good friends and collaborators for over 2 years now. It is true that Xavier was ranked #1 for a very long time. If it wasn't for him taking time away from competitions to work on DataRobot, he may still be ranked #1. Owen, the current #1 ranked Data Scientist, worked with us at Travelers and I've known him and his wife for about 10 years now. Thus, I would say that so far we are recruiting from the intersection of our own network and the Kaggle leaderboard. That said, I think recruiting kaggle-proven data scientists is one of the best ways to ensure you hire a quality data scientist (although there are plenty of other factors that matter like real-world experience building and implementing predictive models).

acoyfellow: There are some startups using Datarobot. We have a cloud product for SMB's and startups as well as an on-premise version for enterprise.

jawns and blauwbilgorgel: Sometimes I use "Data Scientist" and "Statistician" interchangeably, but I do think that stats is a subset of data science (I subscribe to Drew Conways venn diagram definition of a data scientist). I do know plenty of people who call themselves statisticians but actually meet all the criteria for being a data scientist. Regardless of what they call themselves, true data scientists are extremely rare.

If there are any true data scientists (by Drew Conway's definition) that are interested in working with us, please send us an email at careers @ datarobot. We are also looking for great Python, Javascript, and Spark engineers. http://www.datarobot.com/careers/

Seems like DataRobot is building their team directly from Kaggle's leaderboard [1]. Yet another top Kaggle competitor the article doesn't mention who joined DataRobot is Xavier Conort, who was ranked #1 for a very long time.

I'm very curious to see how they go about automating preprocessing and identifying algorithms to use, as I thought this step was part of the hard-to-automate magic that separates your data scientists from data analyst/statistician/etc.?

[1] https://www.kaggle.com/users

I met Jeremy and the DataRobot team at Techstars Cloud (San Antonio, TX), where I was a mentor. You didn't need to be a genius to know that they were going to do well. Congrats for this series A!

Are any startups using DataRobot for any interesting features? I would love to hear about some real world examples :)

We have worked with DataRobot for quite some time now. Our company is called Next Glass and we're building a really cool app that takes the guesswork out of choosing wine and beer by analyzing their chemical properties. Our business model is built around data science and we chose DataRobot to power the predictions that drive our app. Both the software and the team have been amazing to work with. It's clear that they deeply care about the success of their customers. I highly recommend them regardless of whether you are a novice or expert in data science!

In this article, I count eight instances of "data scientist" and zero of "statistician." Three instances of "data science" and only one of "statistics."

I suppose "data science" is a slightly broader term than statistics, but isn't statistics still at the heart of it all?

Edit: Here's a WSJ blog post arguing that "data science" and "statistics"/"data analysis" are separate concepts: http://blogs.wsj.com/cio/2014/05/02/why-do-we-need-data-scie...

I'm still not sure I'm sold on the idea that statistics is a subset of data science and not the other way around.

Most companies are still struggling with the infrastructure to collect and analyse data. This is largely an exercise in distributed systems, and definitely not expertise you'd find in the statistics dept.

Data scientist is the sexy term for statistician. It's almost as overused and undefined as "big data".

You could also call an all-round data scientist a unicorn, since they do not seem to exist. Data science requires a combination of maths/stats, computer science and domain expertise. A statistician who does not know how to run a regression on her dataset, or works on clicklog data without understanding CTR and ROI in a business context, is not really a data scientist.

Here is a Venn diagram: http://www.itworld.com/sites/default/files/data-science-venn...

To DataRobot: Grats on the funding and attracting top talent! Awaiting your product when you release in the fall.

Reply on news.ycombinator.com