End meeting madness with Frejun, which listens and takes down minutes as you speak


Hyderabad-based bootstrapped startup FreJun’s app lets a user set meeting agendas, collect minutes with voice-to-text technology, and action items. 

Like ’em or hate ’em, meetings are a must. They help foster creativity, innovation, collaboration, and ensure proper information exchange. But they’re often chaotic and time-consuming – HBR says employees spend almost 23 hours a week in meetings. How does one ensure that the meeting momentum doesn’t fizzle out? Hyderabad-based FreJun aims to make businesses and teams more efficient, effective, and collaborative. One meeting at a time! 

Subhash Kalluri, who was working for a nanotechnology startup and worked on B2B sales, felt there was a strong need to make meetings more efficient and effective. “While there is someone who can take and record the minutes of your meeting, it is a manual process,” Subhash says. He adds that there are products that listen to conversations and summarise them, but “taking minutes of the meeting is different”. 

He roped in friends Sujith Kumar, Lenin Dubasi and Ravi Chandra Mokanala to help him in this endeavour. 

FreJun was started early this year as a SaaS platform, a web app and a mobile app that allows users to set meeting agenda, collect minutes of the meeting (by converting speech to text), and action items. The functions sync with more than 100 work tools to provide instant access to all secured files in one location. The bootstrapped startup is incubated at IIIT’s Centre for Innovation and Entrepreneurship (CIE). 

How does it work?

The user and attendees of the meeting get a unique web page link 24 hours before the meeting; this is where they can add an agenda. The agenda is shared with all meeting attendees two hours before the meeting, giving them enough time to prepare. 

The same webpage is used for collecting minutes and action items during the meeting. After the meeting, the summary of the meeting, agenda, minutes, and action items are shared with all attendees. 

“The user doesn’t have to worry about making sure that the agenda is seen by all the attendees or minutes are shared with everyone. Everything is automated with a minimum input from the user’s end,” Subhash says. 

The user can use two input methods, typing or voice. Using voice, the user can give voiceover commands as well as take notes. FreJun doesn’t listen to the user’s conversations actively; it starts only when the user wants it to do so. 

Once the voice-to-text feature is turned on, FreJun listens to the conversation (without recording). It takes a “wake” word – “set Agenda” or “take notes” for it to start work. 

Also read: 14 Hyderabad-based deep tech startups to watch out for this year

READ  Motorola reportedly plans a Baidu smart speaker for China

Why voice to text? 

When the team started its research and launched the product, the main feedback from beta users was the same: the biggest issue with minutes of the meetings was stopping the discussion midway to take notes.

Subhash says it took the FreJun team four to five months to build the product from scratch. The first couple of months went into trying out various tech libraries to see which library offered better accuracy and less latency. Tech libraries group similar programmes and codes for better computing. 

With FreJun, the idea was to bring different modulations and commands that come with speech into text. So, the team worked on different kinds of libraries that would match their need and requirements.

Subhash and Sujith, Co-founders FreJun

Once they found the right library, things started falling into place. Subhash says FreJun has developed its own set of libraries for some of the voice-to-text commands. 

“Luckily, we got some external help from IIIT as part of the SSAD project coursework; a couple of students helped us in development. This accelerated the entire process for us. We have the basic foundation to build a complete hands-free, voice-based meeting tool for the next generation,” Subhash says. 

The market and revenue 

According to a report by Grand View Research Inc, the global speech and voice recognition market is estimated to reach $31.82 billion by 2025. Over the past year, apart from biggies like Google and Facebook looking closely at voice-to-text, multiple startups, including 2018 Tech30 company Slang Labs, are also looking at voice-based technology. 

FreJun, which is currently focused on the voice-to-text model, works on a “pay-as-you-go” SaaS model. As a user, you can use FreJun for free for six meetings in a month. The team starts charging $7 per month once the user crosses the monthly limit. A user can also opt for the $70 yearly plan. 

Some of FreJun’s clients include companies like Way2Online, Latitude Technolabs, and Linuxweb. The team claims to be making between Rs 10,000 and Rs 20,000 per month as revenue and intends to grow at least 20 to 25 percent in the next few months.

READ  All you’ve ever wanted to know about Blockchain and what it can do for your business, find out at the YourStory Blockchain Developer Meetup

“We want to be a complete voice-based tool for managing not only minutes of the meeting, but end-to-end meetings. We also want to work on work tool integrations. We are positive that we can make a significant change on how people work in the next three to five years,” Subhash says. 

Source: Yourstory

To Read Our Daily News Updates, Please Visit Inventiva Or Subscribe Our Newsletter & Push.



Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.