A new startup from MIT and Brown lets users transform, visualize, and model data through a graphical user interface.
In a recent post we described the use of AI and machine learning to expand the user base of BI tools, and enable users to perform increasingly more sophisticated analytics. The objective is to reduce time to insights (TTI) by enabling users to uncover patterns, and conduct accurate forecasts, root-cause analysis, and simulations.
A new startup called Einblick is a welcome addition to the suite of tools we listed in our previous post. What really sets Einblick apart is that it lets analysts and managers perform interactive analytics and data science through a slick and engaging user interface. As the video below demonstrates, it empowers non-technical users to participate in data-driven discussions and analytic projects:
Einblick is based on the Northstar project out of MIT/Brown. If it looks familiar that’s because we featured Northstar at two 2019 conferences that I chaired: Strata Data (San Francisco) and the AI Conference (Beijing). Northstar was still an academic project in 2019 but it quickly became a crowd favorite at both conferences. In watching people interact with Northstar, I was amazed how quickly people learned the interface while playing with toy datasets that came with the demo. I can imagine how compelling Einblick can be once teams start using it on datasets they really care about.
But a slick UI is only a means to an end. As we noted in our previous post, what companies need are tools that reduce their TTI, increase their productivity, open up advanced techniques, and help them exceed their KPIs. Einblick definitely helps them achieve those objectives.
Einblick lets teams use sophisticated data analytics to support many forms of decision making and strategic planning:
- Descriptive analytics: Use data to understand what happened in the past. Traditional BI tools (charts, dashboards, interactive analytics) are mostly used for this type of analysis. With that said there are a new set of BI tools (like Sisu) that use machine learning to assist analysts as they wade through high-dimensional data sets. These new tools surface key drivers, uncover patterns, and even suggest charts. Besides providing a highly interactive interface for creating data visualizations, Einblick is able to automatically surface patterns and key drivers. But if you want to track KPIs in real-time, you will need a monitoring system that automatically refreshes data and supplies alerts.
- Predictive analytics: Use data to build predictive models. Forecasting and churn models are common examples in this category. But aren’t there already (autoML) tools that let non-technical users build models? Such tools do exist – think of KNIME, Rapid Miner, Alteryx – but many of them do so by mimicking workflow engines: data comes in, you perform some operation, and the output is passed to another operator. I’ve long wondered whether a workflow-like UI is optimal, and after playing with early incarnations of Einblick I’m convinced that their user interface is a better fit for non-technical users. Einblick enables users to build and share predictive models, and also combine and refine multiple data sets. More importantly, Einblick users build models and data apps gradually, using an engaging interface that lets them combine visualizations, models, and data analysis.
- Prescriptive analytics: Use data to build what-if scenarios or build simulations. You can use Einblick to understand the importance of key variables and predictors, create and evaluate scenarios, and soon Einblick will include advanced tools like Monte Carlo simulation.
You can use Einblick to analyze structured data and unstructured text. The interactive interface lets teams of technical and non-technical users collaborate on projects. Users who are comfortable with Python can even add functions that other team members can access as operators through Einblick’s compelling user interface.
The company is just getting started, today Einblick announced $6 million in seed funding from lead investor Amplify Partners with participation from Flybridge and Samsung Next. Einblick and other startups are going to continue to make analytics and AI accessible to more users. Over the next year I am expecting more innovation in BI, data analytics, and AI tools aimed at analysts and non-technical users.
A trip down memory lane … Project Northstar, circa 2019:
Northstar Demo @northstar_ds from @MIT_CSAIL & @BrownCSDept at #StrataData Big thanks to @MSRDatabase @erichorvitz for sponsoring the Surface Hub, @bigdata for inviting us, and @GoogleAI @JeffDean @edchi @intel @DARPA @MSFTResearch and many others for funding this project. pic.twitter.com/02k5x2gouC
— Tim Kraska (@tim_kraska) March 27, 2019
— Ben Lorica 罗瑞卡 (@bigdata) June 20, 2019