VoiceBase, Inc
VoiceBase BigVoice: How to add your CustomerVoice-Channels to your Big Data efforts

There are 4000 times more words spoken by a company's clients in its call centers and daily interactions, than there is typed on all of social media about the same company. Companies spend millions to monitor, track, and manage social media to learn from mostly strangers; meanwhile information from direct clients with problems, issues, desires and needs are, for the most part, ignored. Imagine being able to spot hot leads, angry clients, potential churn, fraud and everything else, in a fully automated way, with high speed and low cost-directly from your client interactions.

Walter Bachtiger, CEO and Founder, VoiceBase, together with his team created a company that enables just that; access to any spoken information, in a fully cloudbased, low cost way, to automate business processes and seek insights into all spoken communications of a company. With simple APIs, customers can now extract actionable intelligence, quotes, and important information from voice recordings.

“Our uniqueness lies in our continuous effort to develop solutions that solve a key problem for our customers. Whether that is call redaction for call centers, global search for educational lectures, auto notes for conference calls, or simply an accurate and low cost transcript for market researchers,” notes Bachtiger. “Every industry that creates and manages massive amounts of audio or video recordings can benefit from our solution, and it is up to us to provide them with innovative tools to increase the value of their spoken information.”

As a horizontal platform, VoiceBase serves a wide variety of vertical use cases; NASDAQ, who is having financial earnings calls searchable, HireVue who automates part of the job candidate evaluation process by flagging certain words and phrases spoken in interviews, and Twilio’s customers who are able to get automatic transcripts, and extracted keywords and topics for all of their recorded calls.

“We have been working towards providing an extremely accurate machine transcription, at a disruptively low cost, to businesses of any size. With a more accurate transcription, the data and actionable insights extracted hold even more value,” says Bachtiger. Third party tests have shown that VoiceBase is the industry leader in automated speech recognition (ASR) overall accuracy.

Every industry that creates and manages massive amounts of audio or video recordings can benefit from our solution, and it is up to us to provide them with innovative tools to increase the value of their spoken information

This is extremely important when data mining techniques are applied to the transcript to detect trends and enable predictive analytics such as PCI detection, automatic call scoring, and customer interaction identifiers like ‘Hot Lead’, ‘NonProspect’ or ‘Potential Cancelation.’

In today’s world of technological advancements, speed is always a factor. Bachtiger’s team has been focused on supplying these actionable insights in a competitively fast manner. By using mapreduce technologies to parallel process snippets of a single recording VoiceBase can return this valuable data within minutes after the call.

As a competitive swimmer growing up, Walter learned early that honing and perfecting each facet of his sport from the start, the turns, the stroke, cadence, breathing all worked in combination with each other to win the race. At VoiceBase he now applies the same strategy, “We want to be the best, not just at one key component of an automated speech processing solution–but every component, including a highly scalable platform”, states Bachtiger. “We employ unique hiring approaches and aggressive investments in R&D to achieve that. Our non-existing churn indicates that we are on the right track.”

With the evolving digital world, “We focus on replacing the traditional speechto-text systems with automated and efficient speech engines, across multiple verticals worldwide, for effectively turning voice recordings into a searchable database,” concludes Bachtiger.