Companies like Amazon, Google and Apple are mostly focused on voice-command assistants like Alexa or Siri, with the bet that voice interaction will become the next frontier. An alternative is voice technology that takes a recording of people talking and turns it into text.
With Voicera the technology enables those attending businesses meetings to focus on being present and to collaborate, rather than worrying about taking notes and remembering key take-ways and action items. Voicera allows provides voice AI and meeting assistant functions for in-person and phone conversations, automatically taking notes and actions for the user.
The company’s “Voice AI,” technology is called EVA (Enterprise Virtual Assistant), takes care of tracking conversations. To discover more, Digital Journal spoke with Omar Tawakol is the founder and CEO of Voicera.
Digital Journal: Why are so many tech companies investing in voice technology?
Omar Tawakol: Think back to the first time you saw Apple’s demo of a GUI. Once you saw how the world was going to change you couldn’t unsee it. The world had to evolve to an intuitive consistent interface that made it easier to relate to machines.
This is the same, except for instead of asking us to relate to a machine interface, it is time for machines to respond to us in a way that is natural to us. That is what is driving voice. Human speech is simple, loosely structured and common. Language is a social contract where everyone learns a language with the understanding that if they just speak others will understand them. Now that contract is extending to machines not just other people. The wave of investments in voice is a recognition that there will now be a tectonic shift in the way computers interact. Who wouldn’t want to get ahead of that?
DJ: What are the benefits of voice to text technology?
Tawakol: Once you digitize voice into text, you can start to leverage all the natural language processing we have built up over the years. We have systems for document management, workflow, sentiment analysis, analytics, CRM etc. All these systems start with text as the input, not voice. Good speech to text technology generates the content that will now be the new inputs for these systems.
DJ: Which sectors use this technology?
Tawakol: We have customers in almost every sector. Banking, technology, auto, transportation, communications, education and consulting. Our technology is about voice collaboration so like other collaboration capabilities we are widely used. The more interesting question is – who isn’t using voice.
DJ: How accurate is voice to text technology?
Tawakol: It really depends on the scenario. In simple, single speaker, constrained language applications – voice is amazingly accurate. In multi-speaker, noisy environments with broad language model expectations, the hype I believe exceeds the reality. Our application of voice is focused on multi-speaker meetings. We also take the notes out of that scenario and attempt to drive workflows like CRM and project management. That requires the highest levels of accuracy. We have built several techniques to produce the highest levels of accuracy for these more challenging applications.
DJ: Why did you develop Voicera?
Tawakol: We want people to be way more productive. People spend the majority of their time going from meeting to meeting without a clear idea of everything they agreed to. So much time is wasted because of that. The flip side of that is the avid note taker that has their eyes on the computer in a meeting while they should instead be looking and interacting with others. In this sad dysfunctional state of affairs that dominates most of our workday we thought there could be a better solution. So we created EVA the enterprise voice assistant that takes notes and updates workflow for you. The result has been very encouraging.
DJ: How have you ensured Voicera is different?
Tawakol: Our whole premise has been counter most of the competitors approaching the space. We don’t believe that people go to five hours of meetings to then read five hours of transcripts. Who has that kind of time? Instead we focus on creating the actionable highlights and summaries that will motivate follow-up and action. That requires a different set of techniques and a higher accuracy.
We do create the transcript – but we hope that you don’t have to wade through the whole thing to get the nuggets. We are also ubiquitous in our voice inputs. We aren’t trying to create a conference system or a call app – we instead want to embrace any conversation and pull it into our conversations inbox. Every conversation should be searchable, connected to workflow and available to be shared. We focused on making that simple.
DJ: How have you gone about attracting investors?
Tawakol: My last company had a successful exit. We were purchased by Oracle we started the Oracle Data Cloud. That experience allowed me to recruit a killer team that had great experience from Facebook, Google, Microsoft, Cisco and LinkedIn. When you combine a great team and a compelling market you have a formula that attracts investors. But don’t think that makes the rest easy. Building a great company is neither easy nor formulaic.
DJ: How are you marketing Voicera?
Tawakol: At Voicera we have a very strong marketing leader, Cory Treffiletti – and he is a huge believer in building a product that creates fans. We are also believers in building a product that is naturally viral. Finally, we are building a set of content that focuses on helping people become better, more attentive people. The content is about improving how productive we are and not just about our product.
