Bringing analytics in hands of leaders: Natural Language Query in Python
Kanav Hasija (~kanav) |
Data proliferation is putting pressures on business leaders to become data-driven. Although, leaders have to rely on data analysts to run those queries and get insights out from data warehouses. Its a common principle-agent problem wherein data analysts only ask questions from data which they are directed to ask, but its never a one-way street. One has to flirt with data for a long time to get to know it and leaders get stuck in the loop of data analyst direction as leaders are not equipped with or don't have time to write SQL queries.
This calls for a natural language query wherein a business leader can ask a question in simple plain English and data is spitting out either in a table or graph. This session is guided towards how Innovaccer has solved this problem and provides an architecture, knowledge base building, and natural language processing guidance to build one on your own. The session will also emphasize on the fact that accuracy of such a software will be very poor if it is industry agnostic as SalesForce and ThoughtSpot have tried in the past. Thus, one has to tame it to their own business context or vertical.
Basics knowledge on natural language processing, not even how to code it, but what are its basic components.
Kanav Hasija is Co-Founder and Chief Product Officer at Innovaccer. He has developed a healthcare data platform with his team which helps connect to various healthcare IT systems to get a longitudinal view of the patient record and turn it into analytical insights on risk, cost, and utilization behaviour of patient to act on them and treat them before they get sick to reduce the cost of healthcare. The platform today has more than 10 million lives on the platform and an estimated $1 Billion has been saved till date in US healthcare costs while keeping people healthy with a quality of care bump of 15%.
He is a coder and mathematics enthusiast since the age of 10, completed his bachelor in engineering from IIT Kharagpur and pursued higher studies in Intellectual Property Law from UNH Law in the US. He is recipient of various awards like Samsung-Stanford Patent Prize, Honorable Mention for Excellence in Technology, Best Graduate Student Award, and is also an author in a few publications like IEEE.
Harshil Rastogi is a software development engineer at Innovaccer. He has worked on various enterprise-grade software components in the fields of data management, data transformation, and natural language processing.