By Deena Douara
“People don’t like waiting on a call,” says entrepreneur Giri Amarakone. More importantly, they don’t like waiting for incorrect information.
At its core, this is the genius behind deepPiXEL, which develops artificial intelligence bots for companies.
The AI startup is beginning its venture by tackling one of the most common hassles we all face: accessing information from customer service effectively and efficiently.
“We are leveraging AI as a first step in creating smarter call centres,” says Amarakone, who returned from a stint in Silicon Valley to launch deepPiXEL here in Toronto.
The company connects brands to customers online, on their phones and on social media, through a messaging service that can be accessed through common platforms like Facebook, Slack or Twitter – any platform with an open API.
A user simply types in their question – “How do I make a return?” “How much is this product?” “Do you have it in stock?” – and the service responds based on programmed responses, algorithms and machine learning. It could also handle purchases, returns and order tracking.
“We envision a world where almost every business will be on some sort of messaging interface,” says Amarakone. “So while you’re sitting on your TTC ride home you’ll be able to do a lot of things like pay bills and set up meetings.”
He says deepPiXEL distinguishes itself from the competition by having designed a “purpose built” next-generation AI system for language processing which is highly discriminative. The system partitions and secures customer data from the likes of Google and Facebook.
It’s clear that companies can save money by automating some elements of customer service but don’t people prefer to speak to other people?
“If you can get to the right human, yes, but getting to the right human is difficult,” says Amarakone. “There are systems in place to triage before taking you to the right person.”
He explains that deepPiXEL’s bots replace the first tier of customer support where simple questions that don’t offer any problem-solving opportunities get asked. Even those responses can be off-brand, or incorrect when policies change, for example.
“It takes valuable human capital to address those concerns. It’s not a thing that humans like to do. It’s very repetitive.”
While customer service is an industry concern, more important, Amarakone found, were marketing efforts. He says deepPiXEL works on that front as well, offering proactive direct messaging options that are far more personalized than retailers’ email blasts, based on past messages and platform-provided insights.
Though deepPiXEL launched with a retail focus, Amarakone describes applications that go far beyond, including opportunities for HR departments to field numerous questions and to aid with training and information retention. Finance is another major area of focus, with the sector desperate for innovation to face competition from fintech startups.
The company has competition, to be sure – indirect competition like Google, or web developers who are less specialized – but Amarakone feels confident and “bullish” that they are in the right time and place to succeed.
“The main reason (to work in Toronto) is that we need talented AI experts and engineers, who are readily available because of the University of Toronto and Waterloo and other universities, and also due to immigration policies that bring in the brightest minds.”
Amarakone, who immigrated to Canada as a youth, believes strongly in hiring immigrants, and the team hails from across the world – Nigeria, Korea, Azerbiajan, China, Malaysia, India and Canada.
“We take pride in finding people who were marginalized in the Canadian system,” he explains. “We find they’re more eager to make a difference. As a startup, that hunger is what drives us.”
“Innovation is happening here in this place,” says Amarakone, who received support from Founder Institute prior to joining the University of Toronto’s DCSIL incubator and the IBM Innovation Incubator.
“If you can get funding here and get a customer base here, you can out-innovate even Silicon Valley.”