Connect with gaming and metaverse leaders online at GamesBeat Summit: Into the Metaverse 3 this February 1-2. Register here.
As call volumes skyrocketed during the pandemic, contact centers turned to AI to help distribute the workload. But even before the pandemic, customer service departments were experimenting with automation solutions, including chatbots and transcribers, to streamline operations. A 2019 Deloitte survey found that 76% of contact centers were planning to invest in AI in the next two years. According to that same survey, 57% of companies were testing the use of AI in assisting customer service agents.
Anticipating the trend, Karan Kashyap founded Posh Technologies, a Boston, Massachusetts-based conversational AI and natural language processing technology development company, in 2018. Today, Posh announced that it raised $27.5 million in series A funding led by Canapi Ventures. Kashyap, who serves as CEO, says that the proceeds will be put toward supporting additional investment in product research and development and the expansion of Posh’s platform.
“Posh’s growth accelerated during the pandemic amid the increasingly digital world which we continue to live in. Just as the pandemic started, we were already getting ready to hit the gas pedal. The pain points and needs of financial institutions changed from the pandemic to our benefit, including needing to better manage customer service on a 24/7 basis, managing increased call volumes from closed branches, and doubling down on self service solutions,” Kashyap told VentureBeat via email. “There was also high turnover for those customer service jobs — the ‘great resignation’ took a toll on call center jobs too. While Posh’s aim is not to replace human agents, our technology helps our customers address higher volumes and augment their current service models.”
Augmenting customer service
Kashyap, who has a bachelor’s degree in computer science and a master’s in AI, developed Posh’s technology while studying at MIT. The platform provides chatbots that automate customer questions and workflows on the web, SMS, and messaging apps for tasks like checking hours and making payments. A separate IVR bot replaces traditional dialpad menus with natural, voice-driven conversations with customers.
Intelligent Security Summit
Learn the critical role of AI & ML in cybersecurity and industry specific case studies on December 8. Register for your free pass today.
On the backend, Posh automates contact center and help desk FAQs and workflows, leveraging machine learning and natural language processing to give chatbots “memory persistence.” Concretely, Posh’s systems train on domain-specific data so that its chatbots understand some of the nuances of a given industry’s — and company’s — language.
Posh integrates with live chats as well as other “API-friendly systems” (e.g., digital banking databases and telephony) and escalates to human reps if need be. Customers get metrics showing how conversations went and where areas for improvement might exist.
“Our AI can easily manage routine inquiries without requiring staff involvement. We see it as the first line of defense to get people out of queues while also enabling round-the-clock self service,” Kashyap said. “Credit unions and banks are often able to answer customers’ questions directly on their website through the Posh chatbot feature. In cases where the chatbot doesn’t have the right answer, it can intelligently escalate the request to a call center or in-person representative, significantly improving both the amount of money spent on customer service as well as the customer experience.”
Beyond incumbents like Google, Microsoft, Salesforce, and Amazon, Posh competes with a number of startups in the expanding call center automation space. Yellow.ai, a chatbot platform headquartered in Bangalore, India, recently raised $78 million in venture capital to expand its platform globally. There’s also Ada, a Toronto-based startup developing AI-imbued customer service chatbots.
Grand View Research anticipates that the global contact center software market will be worth $90.6 billion by 2028, if the current trend holds.
Kashyap argues that Posh’s focus on the financial services industry gives it an advantage over rivals targeting a broader range of segments. To date, Posh has partnered with more than 50 financial institutions to deploy web-based and mobile-based digital agents, and the company’s software handles tens of thousands of chats per day and reaches over 5.5 million people.
“We serve approximately 50 community financial institutions — banks and credit unions — across the U.S. and their end users and members. Our digital assistants and voice banking assistants handle tens of thousands of requests a day on behalf of these financial institutions,” Kashyap said. “We are very focused on financial services and thus train our AI models to be very domain-focused. Not only are we focused on training models with the goal of automating routine banking inquiries and workflows, we’re also using AI to glean insights from conversations that pass through our system — for example, uncovering operational root causes or detecting anomalies.”
Going forward, Canapi Ventures partner Neil Underwood expects that 40-employee Posh will benefit from expanded access to credit unions, banks, and prospective talent through its other backers Curql Collective, CMFG Ventures, JAM Fintop, Human Capital, and Piedmont. In the coming months, Posh plans to ramp up hiring to keep pace with what it describes as “surging” demand.
“Beyond answering questions, Posh has developed a competency in helping banks complete simple banking transactions. Especially for credit unions, who are highly focused on member experience, this can be a meaningful value add,” Underwood told VentureBeat via email. “Over time, we anticipate that the Posh platform will be used by credit unions and banks to drive entire banking interactions.”
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.