Skip to content

Student Startup Mines Employee Communications for Business Intelligence

By: Diane Stirling
(315) 443-8975

  Data mining for business intelligence
  Student Ross Lazerowitz

The majority of his time, Ross Lazerowitz is a full-time undergraduate student studying two programs - Systems and Information Science, a dual program between the School of Information Studies (iSchool) and the L.C. Smith College of Engineering and Computer Science at Syracuse University, and Information Management and Technology at the iSchool.

Then, in all the spare time he can muster, he’s forming a startup.

But the company Lazerowitz and his business partner Paul Duan are developing isn’t just any startup. Their ideas are so potentially groundbreaking, they have earned the attention of top national business executives and won best prize in a major technology competition.

Their company, Obelisk, is providing a business software solution that looks at the collective intelligence of companies through standard communication tools and via exchanges taking place between employees—parsing the information contained in emails, calendar items, and chats. Using those transactions as a type of social network, the technology “reconstructs the company’s knowledge graph. It gives us a view of not only what [employees] know, but also who they think [in the organization] knows other information. It provides a graph that illustrates the available skills and how to organize them,” the pair explains.  

The technology application views employee interactions within the context of professional tasks. It finds that employees’ communications also include information about what skills are needed to complete the tasks, how they could be solved, and who workers report to or need to ask help of when dealing with those tasks. “Because we can infer the what, the how, and the who, we have all the information we need to build our algorithms,” Duan said.

Extracting Info from Emails

Using state-of-the-art natural language processing and network analysis algorithms, the technology privately and securely analyzes online interactions and automatically infers employee skill levels and learning goals. An algorithm extracts topics from each conversation, infers a employees’ relative proficiency in each topic, and determines the skills needed most, according to Lazerowitz. This is so different from the systems currently available because those are focused mostly on the output measurements of workers, whereas Obelisk’s technology looks at worker inputs, identifying skills, patterns of work, the availability of solutions, and many more functional assessments.

“Gaining direct access to the company’s collective knowledge allows us to solve previously intractable challenges that were previously the turf of management consultants. We’re both creating a set of products that empower decision makers to make better decisions by providing them with better data and insights, and also by automating some parts of the consulting,” Duan added.  

Prizes and Accolades

Obelisk Technologies won the $10,000 top prize from the Lumina Foundation/The Economist Magazine Quantified Work Startup Competition earlier this year, chosen from hundreds of applicants. The company also has been featured in The Wall Street Journal, Lazerowitz said.

The duo have impressive business development histories. In addition to creating a viral application, Beertext.us, Lazerowitz is co-founder of Blu Arc Media Co. and served last summer as a product security intern at Splunk Inc. While at that San Francisco firm, Ross met Duan, a data scientist who was working in fraud and spam detection machine learning algorithms at Eventbrite.

Both Parties Benefit

Businesses can benefit from the technology by having an interactive dashboard and search tool that shows detailed employee profiles, including skill assessments, skill profiles, a skill gap analysis, and combinations of proficiencies within the workforce that are important to the industry. It also can provide analyses on top performers and offer human resource strategy tips.

Employees will benefit from the technology, too, according to the founders. They receive a skill summary that provides recommendations on topics or skills they may want to learn or add in order to advance in the company or to obtain a different position. Information gleaned from the reports also provides tips on personality traits and how to best utilize or improve them.

Plans For 2014

While continuing on his course of study for the coming semester, Ross is keeping the company’s goals on track, too. He said Obelisk plans to have a demo ready for customers in the new year that will illustrate how some of the technology’s building blocks function. Plans for the second quarter include taking the technology into the marketplace, Lazerowitz said.

Although this software is the firm’s first product, the pair has a vision to build an extendable platform and for the technology to perform wider and more extensive analysis, they say. Other applications possible with the software include team selection for projects, performance analysis, and automatic knowledge base creation.

More information about the company is at  www.obelisktech.com, while news about the company’s award is online at: http://www.luminafoundation.org/newsroom/news_releases/2013-09-30.html

Back to top