Sramana Mitra: One thing you didn’t mention is, the channels through which customers interact with an enterprise have expanded. Each of those has become a source of data. Any kind of messenger chat has become a source of data. People are touching the enterprise through many channels. Each of them is generating data. It gives AI a lot more ammunition.
Krishna Raj Raja: The omnichannel made AI more complex as opposed to simplifying it.
Sramana Mitra: But it gives you ammunition with which to model. Earlier on in the 90s, the channel through which data was collected was phone. Someone was taking call notes. That was about all.
Krishna Raj Raja: In B2B, 90% of the data come from calls and emails, and not WhatsApp. It’s so complex that you have to show it on Zoom. The point I want to highlight is, each channel has its own NLP challenge. I’ll give you an example. Sentiment models are publicly available.
When I started the company, I naively assumed that I was going to use Amazon’s sentiment model applied to support data. That’s what I thought. It didn’t work. It failed because, in B2B, there’re a lot of nuances. Sentiment is a lot more subtle. We have to build a custom model to handle the nuances of email. Email has its own structure. On top of that email, you have signatures and footers.
You have social media. You’re probably going to see more sentiments in emojis than in words. The models that work for one channel don’t work for another. The omnichannel makes this single extraction more complicated. Data now comes in different forms and fashion. The language people use in different channels is very different. In chat, it’s going to be more casual.
Sramana Mitra: What are you doing with Yellow.ai? They are a B2C focused company that leverages the WhatsApp channel.
Krishna Raj Raja: It’s not an overlap because we’re not doing anything in the chat system. For us, it’s a system of record.
Sramana Mitra: Yellow.ai is your customer; not an OEM partner.
Krishna Raj Raja: They are not a customer yet. They will become a customer is my point.
Sramana Mitra: I understand. My last question is where do you see open problems from where you sit?
Krishna Raj Raja: This one is a very tricky question. My vision has always been that every customer interaction is valuable. We talked about omni channel, text, and voice. One thing that has fascinated me is what about the retail industry. If you go to retail, you have conversations with somebody at the checkout counter. Is there a way to record that?
A lot of information is the ether right now. Today, we can only process if it’s captured. That’s one area. Another one is Slack. Customer communication is starting to happen in Slack. The nature of conversation there is one continuous stream. How do you draw ticket boundaries? These are the different issues from this conversation.
Sramana Mitra: That’s a complex problem.
Krishna Raj Raja: It is. It’s also an NLP problem. You need issue boundaries from engineering and tracking perspective and statistics perspective. The third thing is we bounce customers to different numbers and medium. If you want support, talk to support. If you want sales, you talk to a sales person. Everybody has their own system of records. Sales has their own.
Another thought process is, can you route any question coming in to the right person directly? Right now, we are routing from support. Can you bring the right people and swarm to solve the issue? That doesn’t exist today.
Sramana Mitra: Very interesting. Tell us a bit about the company. What stage are you at?
Krishna Raj Raja: We are a Series B company. It was founded in 2016. It was bootstrapped for the first two years. In 2018, we got seed funded. We got Series A in 2019. Last year, we did Series B. We have around 150 employees. We are spread globally. Majority are in the US.
We are working with some really large enterprises. Salesforce is one of them. Palo Alto Networks. We have great brand recognition at this stage. We are considered the market leader in what we do.
Sramana Mitra: You’re based in Silicon Valley?
Krishna Raj Raja: Yes. Headquarters is in San Jose.
Sramana Mitra: Thank you for your time.
This segment is part 4 in the series : Thought Leaders in Big Data: SupportLogic CEO Krishna Raj Raja
1 2 3 4