Loading...
You are here:  Home  >  Data Education  >  BI / Data Science News, Articles, & Education  >  BI / Data Science Articles  >  Current Article

In-Depth Interview: Machine Learning and a Real-Time Data Platform with Cinchapi’s Bill Sengstacken

By   /  July 12, 2017  /  No Comments

data platformOne of the more intriguing start-ups in the world of Real-Time Data Platforms is the Atlanta-based company headed by Jeff Nelson, Cinchapi. According to the company website, the Cinchapi Data Platform can radically reduce data prep and cleanup through the use of Machine Learning and other technologies. The company also refers to their platform as a “DVR for data.”

Evidently, the Cinchapi Data Platform does what it claims, because Cinchapi recently won the Atlanta Startup Battle. The company beat out nearly four hundred other participants to take home the top prize of a $100,000 venture capital investment from contest founders Paul Judge and Allen Nance.

Charging forward with this influx of cash is Jeff Nelson, CEO and founder of Cinchapi. Mr. Nelson previously worked for Palantir, MailChimp, and Ionic Security, and he also co-founded Blavity. He then started the open source Concourse Database Project that the Cinchapi Data Platform is built upon.

To give a better sense of Mr. Nelson’s vision for Cinchapi and what the Cinchapi Data Platform has to offer, DATAVERSITY® spoke with Bill Sengstacken, Cinchapi’s Vice President of Marketing.

DATAVERSITY (DV): Let’s start with a brief overview of what the Cinchapi Data Platform is.

Bill Sengstacken: The Cinchapi Data Platform (CDP) is an intelligent system built to simplify real-time data preparation, exploration and development.

We do this by combining Machine Learning with human intelligence to make it easy to combine multiple data sources into a comprehensive view that reveals hidden insights and facilitates dynamic automation of complex workflows.

The CDP connects to existing databases, APIs, and devices and continuously streams data from those disparate sources to a central repository. Once centralized, the platform uses intelligent processing to find relevant insights that are curated for humans to review during exploratory analytics. As humans interact with the data, the platform automatically learns from them and begins to proactively automate their workflows.

DV: What does “Ask, See, Act” mean at Cinchapi?

Bill Sengstacken: The CDP exposes its functionality through a simple Ask, See, and Act workflow.

First, users ask questions to discover data using conversational queries like, “Which factories have lower than normal output?” As they explore further, users can drill down on results by asking a contextual follow-up question like, “Which ones are in the Midwest?”

Users see results through curated insights and visualizations along with additional contextual information that is automatically generated by the platform’s machine intelligence. The results are powered by the CDP’s DVR for Data, which allows users to rewind live results to see how they looked in the past and press play to watch how they develop over time.

Finally, users can act on results using the platform’s automation framework to continuously run complex workflows in the background as results change. This allows users to easily send real-time notifications and integrate with existing applications on the fly.

DV: At the heart of the platform is ConcourseDB, an open source database managed and maintained by Cinchapi. Can you tell me a bit about what Concourse is and what makes it special?

Bill Sengstacken: Concourse is the world’s only database warehouse designed for ad-hoc analytics of mission critical data across time. Concourse takes the best ideas from relational and NoSQL databases to offer a scalable system for analytics that doesn’t sacrifice strong consistency.

Concourse provides an unparalleled level of rich data processing right out of the box that facilitates the ability of the Cinchapi Data Platform to get contextually smarter over time without explicit training from users.

DV: How does the Cinchapi Data Platform make it possible to work with data from any source?

Bill Sengstacken: Concourse uses a document-graph data model, which makes few assumptions about data and mirrors how humans think about information. In Concourse, data is expressed simply as Things that have descriptive properties and are connected to other Things. And since our technology’s conception of data is so simple, we can easily transform data in any format to match the model.

The Cinchapi Data Platform provides a technology called Sponge that is responsible for connecting to external sources, extracting and transforming the data, and detecting when data in the original source changes so that those changes can be streamed to Concourse. Additionally, the CDP uses Machine Learning to detect how data across sources might be related. All of this happens automatically. You simply connect Sponge to the data source and the platform takes care of the rest.

DV: What types and sizes of businesses could benefit most from the Cinchapi Data Platform? In other words, who is it designed for?

Bill Sengstacken: The Cinchapi Data Platform can provide value to any company seeking to get more out of its data investments. Initially, we’re focused on data analysts at midsized companies that want to take advantage of new real-time data streams, but don’t have the resources to figure out how the data can help them grow the business or solve dynamic problems. The CDP helps by greatly reducing the amount of time that is necessary to understand what is important about data and take action on those insights.

Eventually, we hope to scale the platform for use by big enterprise companies.

DV: Who is the end user within a business that you imagine interacting with the platform?

Bill Sengstacken: The Cinchapi Data Platform’s target user is the data analyst that is continuously looking for new insights to answer dynamic questions about the business. This analyst spends the bulk of her time on exploratory analytics – but up to 80% of that work is repetitive. The CDP makes her more efficient by learning from her as she does her job and then proactively using machine intelligence to automate her workflow so she can keep adapting to focus on new challenges.

DV: Congratulations on winning the Atlanta Startup Battle. What does this win mean for your company?

Bill Sengstacken: In one word, it means validation.

While the investment itself is certainly welcome, it’s more about the belief in what we are building. Allen Nance and Paul Judge have been involved with a number of amazing technical products, so in addition to their investment, they each lend credibility to the company and the product.

The win also helps to open more doors with prospective customers and introduce us to a new network of investors.

DV: What can we expect from future editions of the Cinchapi Data Platform?

Bill Sengstacken: We’re working on adding even more baseline intelligence to the platform in addition to more automation capabilities. Our goal is to work closely with our early customers to be responsive to their feedback and ensure that the platform is meeting their needs.

When Cinchapi won the Atlanta Startup Battle, Allen Nance – one of the founders of the competition – said, “We’re living in a world with more data, and the interaction data is more and more complex. Cinchapi is trying to make it human again.” With their focus on the human element and an intuitive workflow, it seems likely that the tech world will be hearing Cinchapi’s name more and more often over the coming months and years.

You might also like...

Using Your Data Corpus to Further Digital Transformation

Read More →