Podcast

Full Episode Transcript:

Jon: Surbhi Rathore, the CEO and co-founder of symbl.ai. A leader in conversation intelligence joins us today on the hello, Human podcast where we discuss the latest topics in artificial intelligence and how it’s being applied in the real world. I’m Jon Knisley, the host of hello, Human and a long-time technologist helping companies adopt and utilize emerging digital solutions.

A big thanks to FortressIQ for sponsoring the program and be sure to hit the subscribe button whenever you listen to podcasts. Whether they are spoken or written, I’m a big believer in the idea that words have meaning. One of my first jobs after college was working for a former newspaper editor so I tend to be careful about what exactly I write or say so I don’t unintentionally convey the wrong meaning.

Since we are talking about conversation intelligence today, this episode promises to be a real eye-opener and I can’t help but feel I’m being judged with every word that comes out of my mouth. Hopefully Surbhi will spare me the hardcore analysis of today’s conversation.

We are fortunate to have a true industry leader to give us her perspective and insight on the exciting news and future of AI around conversation intelligence. It provides businesses with meaningful and actionable insights directly from the voice of customers as well as employees to understand and predict behavior. The technology allows us to measure and optimize marketing performance, detect correct call handling issues, and deliver more personalized and relevant experiences online and over the phone.

I’m obviously excited about today’s conversation and hopefully our listeners are as well. Welcome to the program, Surbhi. Thanks again for joining us on hello, Human and bringing your knowledge and expertise to the program. To start, maybe you can provide the listeners some background of your current venture and journey that got you to Symbl.

It sounds a bit like you’ve been living that tech fairytale, to be honest. A hip startup in Seattle, a Techstars alumni now backed by Amazon. You’ve raised millions in early funding. I think we can make a Netflix series out of it. For those who have gone through it, we know it’s not always rainbows and unicorns. Can you share a little bit about your background and why you started Symbl?

Surbhi: Thanks, Jon, for that awesome intro. Really excited to be here. Started Symbl in 2018 just because I was working in the space around conversational AI before and had firsthand seen the pain and the impact. Both the pain of implementing technology and the impact of the technology implemented in the right way to optimize conversations at scale.

I was really, really enthusiastic to bring to market a developer platform that enables businesses to analyze conversation data at scale, irrespective of whether it is on voice, video, or text channel, but really focusing hard on human-to-human conversations and not chatbots and structured conversations tidbits.

Jon: Last season we did a mini-series on the podcast on women in AI, and what was originally planned to be just two or three episodes to help celebrate International Women’s Day turned into eight episodes because we got such a great response. We touched on everything from optimizing the power of AI for gender equity to applying data-driven solutions for fertility treatments.

You’ve got some incredible recognition including Founder of the Year Award for Women In Voice. You’re one of Amazon’s top 10 women to watch, I think, in 2020. Can you talk a bit more about your experience as a woman in AI and your work to help inspire more women to get into data science?

Surbhi: I started out as a software engineer back in the day almost 13 years ago in India. Lived and worked in India for a long period of time. Even at that time, the number of women in the tech space we’re really less. I was always amongst the lesser number of women on the table, but very fortunate to be a part of the community and also sit with people who really remove the biases of interpretation when there is a woman on the table versus a man.

However, over the last 13 years, I went through different roles from building engineering teams to leading teams. Eventually, I saw the pre-sales world for big enterprises. That’s how I transitioned through different technologies, through different applications, and eventually landed in conversational AI.

It’s interesting when we look at women in AI versus women in tech because that’s even smaller in numbers. Personally, I’m really excited about empowering the community to just make it very transparent that it’s not that difficult. It’s just another technology like how technologies have evolved. It’s just a cold start problem today.

Personally, I’m involved and try to read, try to do however much I can in the women in voice ecosystem, women in AI ecosystem. Trying to make sure that I do my bit as running the company Symbl and give equal opportunities in the space. It’s definitely different, so that’s what I would say.

Jon: It’s great. There’s a good bit of excitement in the voice tech market these days. Microsoft acquired Nuance earlier this year for nearly $20 billion. I think that was Microsoft’s second largest acquisition after LinkedIn. Gong raised $250 million this summer as well.

For those that are not familiar with the space, can you give us a little primer on the market and differentiate between intelligent voice automation from conversation intelligence and Symbl’s approach to the market?

Surbhi: At a 10,000 feet view, conversation data is split, I would say, into two prime categories. There’s a human-to-machine conversation, which is more virtual assistant structured, and has a goal and an end objective for that conversation. Then the other one is human to human piece. In between lies an intersection in between both of these that creates the hybrid nature of that conversation type.

I would say intelligent voice automation is definitely a part of the larger conversation intelligence landscape, but it’s more than that. It’s more than just automating. It’s about learning from the conversation and enabling coaching use cases like Gong does.

It’s about enabling compliance on the fly, and not just in the call center. How do you expand this intelligence to other verticals like the webinars space, the sales space, and healthcare like you rightly mentioned? Nuance has been deep into healthcare since the very beginning and I feel this Microsoft’s acquisition of Nuance enables them to have a stronger foot within the healthcare ecosystem. There is so much conversation that happens in healthcare at this point in time, that consumes manual hours of people in order to do the downstream processes.

At the end of the day, it’s all about connecting unstructured data, adding communication endpoints, and adding the system of engagement. How do you convert it into a structure? Put use cases and push them into the downstream systems, which are the system of records because that’s where data should live.

Conversation is just another type of data that exists today, which is just not capitalized on. That’s why conversation intelligence, by my definition, is a larger landscape of how AI and machine learning is used to understand, comprehend, act, and analyze conversations about irrespective of whether it is human-to-machine conversations or human-to-human conversations, and really convert it into actions, insights, knowledge, which can then be capitalized on as a function of growth for the next generation of businesses.

Voice is so ubiquitous at this point of time that communication products that itself are of lesser and lesser value every day. Our movement or our work in the industry is really about how we use this new voice data as the next mile of data and push the boundaries of the next generation communication experience this data is being built.

That’s where Symbl lies. We are enabling developers and businesses to be able to empower their teams and add in their products intelligence easily with simple APIs without having to build extensive machine learning models or set up investments into data science teams where outcomes are still unpredictable. Give them tools in which they can do it very easily.

Jon: That’s great. That’s really helpful. It’s kind of that common theme of trying to turn the unstructured data that lives in so much of our systems and so much of our day-to-day life and turning that into structured data that you can go tackle, analyze, and look at to make the organization better. I guess the good news is that we’ve come a long way from those automated voice response systems from the old days, but I do miss that ability to just hit the zero button and get transferred to a human. That doesn’t seem to work as often these days and then I’ve got to listen to the whole message.

Years ago, we had IQ to measure intelligence. Then we added EQ, emotional intelligence, to the mix of skills that leaders needed to think about. Do you envision a day when we talk about a leader’s CQ? Some measure of their conversational intelligence given the importance of communications and how we articulate our position on an issue?

Surbhi: Yeah, absolutely. I mentioned this before as well when we talk about the platform evolution and the technology evolution in the conversation AI space that sentiments, emotions are definitely one part of the ecosystem, but we are still missing a lot of data on content. What is the type of content which is being exchanged? And how do you measure that engagement and the information which is being capitalized on? Those aspects become a big part of the overall conversation quotient or conversation experience.

The actions collaborated, the information exchange, the entities that are being talked about, any new form of data which is being transferred, and the knowledge which is being generated. All these factors do not come into consideration when you talk about the emotional quotient. It’s still at a very high level. Getting very deep into the content of the conversation and using that to determine what was the quality of the conversation itself will definitely be an evolving theme in the space.

Jon: I agree with you. I think just because somebody may be empathetic, emotional, and understand somebody else’s needs, if they can’t communicate, if they can’t verbalize it, it’s a big miss. I think this works pretty interesting.

This is our second season of hello, Human, and as I mentioned, this season of the podcast is really trying to focus on applied AI. How do you see Symbl being used within different industries? Can you share maybe one of your favorite use cases for the technology?

Surbhi: When developer platforms come to market, mostly there are end business use cases and vertical applications, which have grown an immense value. For our case, it is companies like Gong and Chorus, which are proving the value in the sales and revenue intelligence space, and companies like Uniphore and Observe, all these proving the value for the contact center space in the customer experience.

I see that Symbl providing an unbundled and a programmable way in order to build your own versions of Gong, Chorus, Observe, and Uniphore in the space that you operate in is like the best thing. Being able to understand, act, and coach from existing conversations what will be the next conversations and attach that to predicting the outcomes of the business is really interesting to me personally.

Whether this coaching is enabled in recruitment to the recruiters, it is enabled for trainers, it is for sales agents, or customer support reps, coaching in itself is really I feel the best way in which we can leverage conversations, understand, and improvise on it.

We’ve seen a lot of businesses trying to add coaching experiences within their existing workflows. If it’s a CRM system, can I add that capability within my workflow? Or if I’m a contact center application, probably I can do that. But also on a wider side of all the HR platforms and all the edtech platforms. There’s a big opportunity to coach the trainer or train the trainer sort of thing.

Those are, I think, some of the most interesting coaching use cases, which have immense value and impact immediately. Apart from that, I think accessibility is also another side of the equation, by which you can enable information to be inclusively shared even in real-time and not just imposed.

We look at live captioning, identifying the question that is being talked about in real-time, letting people know the topics of discussion across different webinar events so that you can hop on to different webinars or different events and sessions without having to think about or waste the next one hour if you’re not getting the right information. Matchmaking people with the right sessions and events based on their engagement in previous sessions.

These are all very interesting. Very early use cases in the market, but a lot of value around that. I would say search, indexing, accessibility, definitely a starting point for several businesses, which can be step one. Then eventually arriving at call tracking, coaching compliance for different verticals is kind of like the next step two, I would say, where you can fully capitalize on the conversation data.

Jon: I’ve got to imagine just the shift to more remote work that we’ve had through the pandemic has been beneficial to your business and organizations. Is that the case?

Surbhi: Yeah, that’s definitely true. I think, overall in the ecosystem, we have fast-paced almost 10 times the adoption of digital communications. Symbl really enables people to improvise on digital communication and capitalize on digital communication data. Having that baseline set is really important for us.

With COVID, we went to that unanticipated scale almost overnight without having to push people to add voice and video in their applications. Eventually, I strongly believe that there’ll be a time that every single B2B SaaS application, even B2C applications will have a voice or video component to it.

Then the next question will be how do we use this data, which is currently getting wasted, in order to add more capabilities to our platform, increase retention, identify upsell opportunities, or even build new products for the growth of the business. That’s the cost, I think, where we are at this point in time.

Personally, with Symbl, I’m really excited that we are adding voice intelligence and enabling conversation intelligence at such a crucial time in the era for the entire workforce where communications are now fully remote, which is impossible to even imagine a couple of years back.

Jon: It’s always good to try to find something good that came out of COVID so I appreciate that insight. Years ago, I commercialized a technology that looked at unstructured data and signals data to really help speed analysis. FortressIQ tackles process data with a similar framework. It’s ultimately a big data problem that we’re solving and the application is to look for the patterns in the data. You’ve previously discussed how voice can enable automation, but when you take a step back, it is another example of this sort of big data problem that we’re trying to get into. Can you share more on this?

Surbhi: Yeah, and this big data is not even being stored at this point in time, which is another problem. You can’t act on data if you don’t have that data and I think that’s the part that we are stuck with right now. People and businesses are not even storing voice data because storing audio files and recordings in itself is a huge problem from a storage standpoint.

Once we have the right way to convert this unstructured voice data in real-time to a structure, which then can be used and stored in large capacities, then only we can think about, okay, how do we apply it in predictive, in search, in actions, and so many other ways? But I think we are still stuck at that early problem right now that storing large volumes of voice data in an easy form is still very difficult. I don’t know if that answers the question, but I think we’re not there yet with voice data.

Jon: No, it’s perfect. This has been great, Surbhi. To wrap up, one final question. You can take this one anywhere you want. What excites you about the future of AI in your industry and beyond?

Surbhi: Whenever I look at all the movies of Iron Man, I’m very fancied by Jarvis. I feel that eventually we all will have our own versions of Jarvis. Personalizing the information that you can capture from any human conversation and making it accessible when you need is one of the fundamental areas by which I’m truly excited about. I think if we can truly comprehend conversation data, it is going to unlock so many new use cases that we cannot even think about at this point in time.

I’m very excited about how conversation, voice, video data will be used and acted upon by the industry. That’s what I’m looking for and working on at this point of time.

Jon: That’s great insight and a great point to end on. To recap today’s conversation with Surbhi Rathore, the CEO and co-founder of Symbl AI, a leader in conversation intelligence. In our multi-channel hyper connected world, customer experience is more important than ever as well as becoming harder to manage and control.

The ability to turn everyday conversations into actionable insights using technology from companies like Symbl is a game-changer for many industries and organizations. It helps uncover the voice of customer insights, provides a coaching tool for new and seasoned reps, helps detecting correct conversation issues, and maybe most importantly—democratizes insights across your organization.

This episode has been part of our second season of hello, Human and a big thanks to Elizabeth Mitelman for spearheading the season. That’s a wrap on today’s show. Thank you, Surbhi, for joining us and FortressIQ for sponsoring. If you enjoyed it, be sure to give us that like on whatever platform you’re listening. I’m Jon Knisley and this has been hello, Human.