The Silicon Review
“VoiceBase extends valuable capabilities familiar to text such as a record of the discussion, sharing, storage, access and search — to voice.” – Walter Bachtiger
The single most important part of any call is what’s said. Unfortunately most businesses only collect the text data related to phone calls; date and time, duration, phone number, location, etc and are missing out on the most valuable portion. Voice communications are the richest and, most authentic source of actionable information. The only problem is that it is spoken, conversational and private, making it extremely difficult to work with compared to how text information has been harnessed using big data analytics. VoiceBase’s Big VoiceTM platform now makes it possible to accurately transcribe, detect and understand intent from what was said.
VoiceBase allows users to instantly search and listen to spoken information in the timeline of recordings. Users can manually upload audio or video – using the VoiceBase mobile app, a web browser, or automate the upload process via the VoiceBase API. Recordings are then stored in the cloud for safe and secure, anywhere access. Once a file has been fully indexed and transcribed, it is ready to be searched, shared or mined for critical data points used in predictive analysis. VoiceBase’s predictive technology uses customer’s pre-tagged call data to train unique models to detect the presence or absence of specific events such as; hot lead, rude agent, close-to-cancelling, first-time caller, or positive/negative comments. For events that are defined by a start and stop time, such as sensitive information like PCI, PII or SSN, VoiceBase is able to redact this content to provide a clean repository for the customer to then search through.
How does VoiceBase Work
The REST API allows customers to point VoiceBase to the respective URL and the recording is retrieved from it. Or they can also post the audio as part of their API request. As an API platform VoiceBase is inherently scalable and built to process massive volumes of audio and video content. The VoiceBase hybrid cloud combines the reliability and security of Amazon Web Services with the cost efficiencies of VoiceBase’s own private cloud in a powerfully orchestrated solution. All of the results for the Transcripts, Analytics and Insights products, along with confidence scores are delivered in JSON format using callback or polling method.
The Speech Analytics Solutions You Always Dreamed Of
VoiceBase has three product categories namely Machine Transcription, Speech Analytics and Predictive Insights.
Machine Transcription – VoiceBase returns a fully time aligned, highly accurate transcript in to TXT, WORD, RTF or SRT format. It automatically provides the confidence score for each word in the transcript. Fully time-stamped words combined with a source URL allows a customer to surface specific terms in all of their stored recordings. VoiceBase’s Web Player SDK provides Browser UI components that extend the capabilities of popular commercial media players. Captions are also generated automatically from ASR transcripts and can be retrieved within minutes.
Speech Analytics – Using proprietary algorithms, the company’s platform automatically identifies and extracts relevant keywords and topics in a recording. This provides a user with immediate click-n-play navigation within a recorded event. Users can delete and add keywords to the generated list to manage the way in which their content is consumed, as well as define custom keywords for the machine to spot. This is ideal for a one-to-many use cases or long recordings like speeches, meetings or lectures.
Predictive Insights – With the predictive insights product, customers can identify any event or call classification that is valuable to their business and may be used to automate a business process. VoiceBase can detect unlimited events within a recording, whether they be defined by the entire call or restriced to an exact start and stop time . Recordings can also be scrubbed for sensitive information such as PCI, SSN, Customer ID, first and last name, profanity, or phone numbers to increase data security. By using pre-tagged customer data VoiceBase can create unique models to identify any events specific to a customer’s use case.
VoiceBase’s customers include companies such as Twilio, Kaltura, INVOCA, FreeConferenceCall.com, Nasdaq, C-SPAN, and more.
“The main components that we were looking for were a secure platform, scalability and accuracy. We landed on VoiceBase and after 3 years we’ve reached a very interesting offering where our customers get an affordable price as well as extreme turnaround times and a secure and accurate platform.”
– Zohar Babin, Managing Director, Kaltura
“We believe VoiceBase Insights has the potential to unveil a deeper, much sought after layer of speech analytics, that up until now has been inaccessible. For marketers who drive calls to local businesses, accurate lead scoring on every call, unhindered by human capacity and cost, is game changing”
– CJ Arseneau, Vice President Marketing, Telmetrics.
The company is headquartered in the Silicon Valley in San Francisco, CA. It also has offices in the Ukraine, Russia and India.
Meet the Leadership Team
Walter Bachtiger, Founder and CEO
Walter is the company’s visionary. Raised in Switzerland, he studied data mining and economics, before following his entrepreneurial dreams to the US.
Jay Blazensky, Co-Founder and CRO
Jay played a key role in establishing RingCentral as the leading hosted communications provider, and has a passion for building trusted business relationships.
Jeff Shukis, VP of Engineering & Technical Operations
Jeff served as VP Development for Oracle, Bridgestream, and Promptu, and has proven skills in software development, cloud computing and operations.