Meet the AI startup that translates everyday graphs into captivating stories

How I Met Your Mother’s Marshall would be so proud.

If you want just-announced Google Home for your house, maybe you’ll want this product for the office.

Consider this – you’re sitting in your company’s monthly employee meeting, where you receive performance reports. You’re a little tired, a lot hungry, and when the reports come around, they’re in graph form. You’re supposed to read them thoroughly, so you glance through them, are pretty sure you’re not getting fired anytime soon, and file the report away somewhere in your email for future reference.

You’re totally going to look at it again, just like you totally read through and fully understood the report, right?

Charts are a great way to make everything look 200 percent more official. Too bad they’re not always as explanatory as they look. For data to be useful, it has be to accessible, which often means breaking down several charts or large amounts of data into layman’s terms. That’s fine, but it can get really time-consuming and/or costly for the people who have to do the interpretation.

You’re totally going to look at it again, just like you totally read through and fully understood the report, right?

Mumbai-based vPhrase seeks to take the interpretation stage out of the process with the artificial intelligence product Phrazor. The software-as-a-service (SaaS) extrapolates insights out of data and communicates it according to a company’s preferences. Founder Neerav Parekh got the idea after working at digital marketing agency Iksula and watching his clients struggle with understanding the data his company was giving to them.

The company was supposed to send its clients reports every week, Neerav explains. It was taking two to three hours to come up with the reports, and then it took even more time to explain the data to the clients. In order to help both sides, Neerav decided to go into the business of making data into narratives and founded the startup in April 2015. He previously co-founded online payroll app eVetan.

Neerav believes that data insights should be understandable even to people outside the industry; they should be interesting and tell their own story. Phrazor, then, is the intermediary that can close the information gap between visual aids and people.

Report on demand

Neerav Parekh, founder of vPhrase

The patent-protected product has to be trained for use, a process that takes on average less than two weeks. Usually, this consists of the client sending over sample reports and the type of analysis desired from data. After vPhrase’s team programs it, the software is ready to go and continues to add onto its original learning template. It’s applicable to several fields and can fit into any role that requires large amounts of data analysis, including financial advising and journalism (AI software already writes articles for AP and Fox).

Clients can either play for the license or per report, which can be anywhere between US$0.15 to US$37, and Neerav says the startup has users in India, Sri Lanka, Thailand, and Indonesia.

Would that mean that vPhrase could extend into the sciences as well? Mumbai-based Arya.ai, for example, is an AI created by Vinay Kumar Sankarapu and Deekshith Marla who have just made the consumer tech Forbes 30 under 30 list for Asia. It’s meant to help crunch large amounts of information to help scientists or other researchers. The goal is feasible, says Neerav, but vPhrase communicates in a more vernacular and less academic style.