5.4.22-Startup-Ecosystem-Einblick

BENEDETTO BURATTI: Good afternoon, everybody. My name is Benedetto Buratti and I'm one of the co-founders and head of data science at Einblick Analytics. I'm in place of Emanuel Zgraggen. That's why you see his name. Unfortunately, Emanuel was sick. So I took his place. Emanuel was-- is sick, but is still with us.

[LAUGHTER]

Is a post-doc, was a post-doc at MIT, under the supervision of Tim Kraska in the database system group. And Tim also is a co-founder of the company. So today I will talk about our interactive science platform. It's called Einblick.

So back when we were doing research, we were always inspired from like the vision of those interactive data science platforms, right? In this video that was published by Microsoft a while ago, you can see domain expert analyzing data in those interactive white canvases. And all the action they're doing there looks really natural, right? And so we were really inspired by this.

But if you think about what data science is today, well, the reality is a little bit different, right? So we all know that data science is complex. We already know that it requires a lot of technical background, coding skill, and so on and so forth. But this essentially is creating a roadblock, essentially is an entry barrier for domain expert, that might not be able to code but they still would like to analyze their data.

So to solve that problem, we created Einblick, and it's a web-based application and allows technical and non-technical people to analyze data on our interactive whiteboard. So in this video that you can see, right now it's deployed on a Surface Hub from Microsoft. So it works on touch devices, but also on laptops, tablets.

And the nice thing is that it's really easy to use. It just runs on any web browser. You can just enter your credentials, start using the platform. So the first innovation was really just the user interface, code optional, that allows technical and non-technical people to analyze the data and have a conversation around data.

The second innovation is in the back end. So I know a lot of you in the audience probably work with large amounts of data, right? So if you want to have collaboration on large data sets, well, you're out of luck, because even to do a simple operation would take quite a bit. So what we designed and created is a progressive computation engine that essentially builds up quick approximation of any job that you send to the back end, so such that you already see immediately a first response in a few seconds, right? So that you can keep the conversation going, you're not stuck in the whole processing.

And then we automatically refine the results on the whole data set. And then finally, we develop a wide variety of algorithms, always according to this incremental and progressive paradigm. So we have a built-in AutoML, so even people that are not able to code, if they use our tool, they can log in on a Canvas, do visualization, data manipulation, even with dirty data, and then be able to build a predictive model without ML.

So why Einblick, with respect to the other commercial offerings? Well, first of all, because as I've already said, alignment is one of the biggest problems, right? So having a web-based application allows you to just share a URL link with your team and have multiple people join a Canvas. And you can start having a conversation about your data.

The second thing is that this collaboration allows rapid prototyping, so if you have a data scientist and domain expert, they need to talk about a specific use case. Now they don't need to exchange emails or jump on an in-out call or team meeting. They can just jump in a Canvas, turn on their camera, and have a discussion real time on the Canvas.

And by the way, if you stop at the booth later on, I can show you in real time what I'm talking about. And finally, the teams actually can now work together and share the result. And you can use the Canvas itself to make a presentation. In terms of concrete use cases, we have a wide variety of clients, most notably in manufacturing.

One of our largest clients use the platform to analyze their logistic parts delivery. So it's a car manufacturer. So in the past, they were just using prior experience to analyze the data, right? And maybe they knew which vendor was complicated, but they were not able to cross-relate data that comes from the vendor or which specific part that part in the car was.

So what they did is to use the platform, connected multiple data sources about vendors, about parts, about builds of their car, and so on and so forth, and manipulated data, built monitoring dashboards, and then finally the end product was a dashboard for the factory planners, such that they can go in in the tool, insert their ID, and actually check with which vendors were high risk of probability of being late.

So if this sounds interesting to you, you can just reach out to visit us at the booth. I will be more than happy to have a demo and show you the Einblick live. Otherwise, like we are looking in general for partnerships. So if you have a use case that you would like to try out, please reach out to us. You can see two emails here on the slide.

And most importantly, we have a freemium tier. So we have both on prem and SAS offerings. So if you just want to try out the tool, you can go on our website, create an account free, just to try it out. And of course, if you want to on prem, we can have a deeper discussion. Thank you very much for your attention.

[APPLAUSE]