DataChat: Effective Data Analytics with NLP

Jignesh Patel, CEO

As data sets are growing and diversifying by the minute, a huge demand has arisen to incorporate them on a sophisticated platform to make it useful. However, people with limited technical knowledge struggle with the overwhelming volume of data and are unable to extract meaningful information. Consequently, they turn toward traditional techniques, without realizing that such methodologies cannot solve complex data sets. Considering the need for a solution that empowers non-technical users, Jignesh Patel, Professor in Computer Science at the University of Wisconsin, and his team devised a conversational platform named DataChat. Unlike the usual programming language, this solution uses engineered natural language to generate valuable insights removing ambiguity from the equation. Since the platform can unambiguously understand the intent in the engineered natural language, the required data analysis and machine learning code is automatically created on-the-fly, essentially behaving like a software data science robot.
A primary element of DataChat’s solution is natural language processing, which ensures that the interpretation of data is accurate. It is essential to engineer a surface where any ambiguity can be removed by design instead of waiting for the user to misinterpret and make errors. The platform effectively assists its users in solving business queries, for example, if the client wants to know the reason behind the difference in healthcare costs across the state of California, a thorough search through the data is conducted. Any errors in the data along with suitable recommendations are consistently communicated to analyze and study the different parts of data. Once the desired information is obtained, the overall process of solving the business problem at hand is automatically stored on the platform, which can be used later for similar problems.

DataChat has procured clients across the healthcare, manufacturing, and finance sectors including Fortune 100 companies. The first class of clients the organization caters to are the ones who deal with complex information stored in data warehouses and spreadsheets. In the past their only way to consume new insights would be to seek help from IT and data science teams, or deal with raw data directly. Now they can talk to their data using DataChat deriving insights that they couldn’t have before. In this mode, DataChat allows business users to self-serve analytics and machine learning.
The second class focuses on companies that serve insights via complex dashboards. With DataChat dashboards collapse into a single conversational interface and business users can naturally navigate and drill through charts in the dashboards. IT can still keep control of what views of data are exposed to users, but users are not overwhelmed with a flood of charts, and IT can now easily infer the pathways through which users navigate through data. In addition analytics driven by advanced machine learning algorithms can easily be added by IT in their business delivery workflows.

With a platform that is capable of generating insights in a reproducible way and performing data functions that range from data cleaning to advanced machine learning in a single unified conversational platform, DataChat has set a new standard for chat technology in the market. Envisioning their future ahead, the company plans to aggressively grow their technical team and build a high-powered sales team focusing on North America next year and world-wide in subsequent years. Additionally, they wish to fundamentally change how organizations realize value from data. “Our focus is to revolutionize the way organizations leverage data by empowering decision makers to discover insights by simply conversing with DataChat, allowing organizations to finally realize the full value of their data,” says Patel.