By Sarah Griffiths | December 2020

Bringing new business to startups and fresh innovation to global customers is Oracle for Startups’ jam.

Copenhagen-based startup BotSupply develops conversational, multilingual AI-based interfaces for customers in a range of industries. As a member of the Oracle for Startups program, the company is taking full advantage of the resources available to deliver innovative, new solutions to its customers. In addition to testing and building its cutting-edge chatbot platform using Oracle Digital Assistant running on Oracle Cloud Infrastructure, it’s now developing an offering on Oracle Autonomous Data Warehouse that further extends its chatbot capabilities into the realm of deep data analysis.

Customers already use BotSupply’s original AI-based platform, available in 100 languages, to design, build, and manage chatbots that use natural language processing to handle HR and customer inquiries. Besides providing helpful, automated—but conversational—responses, these chatbots collect lots of data that businesses want to tap for additional insights. However, that data is often difficult to access and analyze. That’s where Oracle Autonomous Data Warehouse comes in.

BotSupply is using Autonomous Data Warehouse to build a new product called Insights to give customers a new way to derive actionable insights from emails, transcripts, and other data sources.

“Autonomous Data Warehouse has prebuilt, out-of-the-box models and we will build additional AI models on top of the Oracle ones to deliver a new product,” says BotSupply CEO and Cofounder Francesco Stasi. “So we can very quickly and accurately get insights out of this historical data that has, in many cases, gone unused inside companies.”

The Insights tool connects users’ other scattered conversational data sources, such as chat logs, call logs, historical transcripts, and ticketing systems, to the BotSupply platform. The tool then discovers and visualizes trends hidden within that vast collection of data to clearly show user intent. For example, by analyzing the data collected from emails, phone calls, and chatbots, Insights could reveal that 70% of a company’s conversations with customers are about delivery status.

“We want customers to dive one level deeper and understand the why behind the conversations,” says BotSupply Chief Operating Officer Giovanni Toschi. “We're not looking at just reporting how many conversations or how many messages, but extracting new insights and making those insights actionable.”

The insights generated could help a business improve its chatbots, lower operating costs, find new revenue streams, and boost customer experience in a faster and more in-depth way.

Development process so far

BotSupply first built a connector to move data from its chatbot management system into Autonomous Data Warehouse, and then it built layers of machine learning on top of those already in the cloud database.

BotSupply is now working on an Insights proof of concept using data from its current products. “The next phase will be to look at other Autonomous Data Warehouse customers not using our solution and see what kinds of insights they need,” says Toschi. Oracle Autonomous Data Warehouse already offers visualization tools, which BotSupply can leverage “to show insights in a meaningful way,” he says.

Data security and more

BotSupply works with customers, such as hospitals and insurance companies, for which information security is a top priority. Having Autonomous Data Warehouse at the core of its product, “is basically mandatory,” Stasi says, to get the attention of big enterprise accounts. They also demand a platform that can analyze large amounts of data quickly, he says.

BotSupply recognizes the benefits of automating customers’ time-consuming tasks. “The least human involvement, especially when you are talking about data and privacy—that’s a great thing,” Stasi says. “The Autonomous Database takes databases to a new level. It’s the future.”

Another big benefit of using Autonomous Data Warehouse and having membership in Oracle’s startup program has been the ability to build on top of the software and partner with Oracle. “It not only makes your products stronger, but it also allows startups and Oracle to co-innovate and co-sell,” Stasi says. “That’s huge for startups.”

All Oracle for Startups members have access to Oracle Cloud experts, salespeople, business development pros, and other mentors to get ongoing feedback, so they can feel confident that their products work well within the Oracle ecosystem.

The BotSupply team is in talks with Oracle about early testing of its Insights product to determine which Oracle customers might be interested in piloting it.

“The chatbot industry will continue to evolve quickly, and so we have to stay ahead of innovation,” Stasi says. “We are always looking for ways to deliver more value to our customers and working with Oracle and Autonomous Data Warehouse will allow us to do that.”

Resources for startups

Members of the Oracle for Startups program receive perks that range from free cloud credits to introductions to enterprise customers.

Photography: Getty Images/Shannon Fagan; Illustration: Wes Rowell and Oracle