UnitQ’s new AI assistant surfaces customer insights from across channels

As customer expectations continue to rise, enterprises are under increasing pressure to provide seamless support experiences. However, keeping up with growing demand can strain even the largest support organizations. 

This week, the five-year-old startup UnitQ announced a new chatbot, UnitQ GPT, powered by a combination of OpenAI’s GPT-3.5 Turbo model and proprietary machine learning (ML) models for analysis, that provides any employee access to insights without requiring data science expertise.

“We’re not just now democratizing access to data, but also democratizing access to insights,” said Christian Wiklund, UnitQ co-founder and CEO, in a call with VentureBeat.

Wiklund described how UnitQ’s platform aggregates and analyzes vast amounts of user feedback data — gathered from app reviews, support tickets, online posts and product engagement data. The company’s machine learning models sift through this information and categorize it into granular topics or “buckets”. 

Additionally, the models extract important metadata about each piece of data, gaining an understanding of elements like the customer’s platform or device.

This thorough classification and analysis provides relevant details into each unique data point that comprise customers’ experiences with a company’s products and services.

“And then you can start asking questions,” he added. UnitQ will respond with answers and insights, even producing graphs of different user interactions and responses showing how frequently they occur.

Since 2018, UnitQ has raised $41 million from backers like Accel and Gradient Ventures to develop a platform that allows product and engineering teams to measure app quality in real-time using conversational AI. 

Prominent customers like Shopify, Pinterest, Uber and HelloFresh are utilizing UnitQ’s solution to gain these insights. Wiklund detailed how early adopters have benefited from UnitQ’s platform.

“By prioritizing quality improvements, businesses can increase retention, conversion and ultimately the bottom line,” Wiklund said.

The power of personalized, scalable support

This attention to the data analysis allowed product managers to inquire about the top issues impacting key metrics based on feedback. Engineers might get suggestions on stability bugs to prioritize. And support agents could gain a deeper understanding of common customer problems. All without needing data science expertise.

To ensure the chatbot provides reliable guidance tailored to each business, UnitQ trains it on their customers’ unique classified feedback datasets. 

“The trick here is like for every customer we have, we have like a data labeling team that will then go in and actually label data, create training data, and then train these models,” Wiklund explained.

This personalized training approach is critical, as no two customer bases are exactly alike. By leveraging hundreds of conversations with a company’s own support agents and customer comments, UnitQ can build chatbots that offer relevant information contextualized by the nuances of that organization’s customers and products.

The results speak for themselves. Pinterest, one of UnitQ’s early customers, now has over 800 employees actively using the platform. “It’s from engineering to product to support teams to research teams, to leadership because everyone has a vested interest in understanding the user base,” Wiklund noted.

Future-proofed, enhanced customer experience

Bumble had seen great success in leveraging AI and machine learning technologies through their partnership with UnitQ. As a leading dating app, Bumble had access to a vast amount of discussion and feedback from users online. However, sifting through this “raw and unorganized” data to gain meaningful details could be a challenge. 

Cathleen Doorenbosch, Bumble’s vice president Customer Care, elaborated on how UnitQ had helped the company do just that. She explained that UnitQ had “allowed us to really listen, like truly listen, and most importantly, really learn from our members and improve everything from our process.” 

By capturing user sentiment and feedback at scale, UnitQ gave Bumble deep insights into what their members wanted and didn’t want. This in turn allowed Bumble to continuously enhance its products, policies, and overall user experience.

Doorenbosch also highlighted how UnitQ had developed new features specifically for Bumble’s needs. Given the sensitive nature of some user reports on a dating app, privacy and anonymity were paramount. UnitQ created functionality to “really restrict access to sensitive conversations” and better manage who at Bumble had visibility into what types of support cases. This showed Doorenbosch how UnitQ was a true partner in adapting their solutions to meet Bumble’s unique requirements.

By future-proofing support with conversational AI, enterprises can scale their operations, enhance customer satisfaction and gain a competitive edge. UnitQ is leading the charge in revolutionizing how businesses leverage the latest technologies to reimagine the customer experience.

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.