Head over to our on-demand library to view sessions from VB Transform 2023. Register Here
Stravito, an enterprise insights company, is introducing proprietary generative AI capabilities on its platform to streamline enterprise knowledge management. The aim is to optimize the search experience by delivering transparent answers, complemented by clickable links to source documents for fact-checking.
While Dropbox and Box have recently introduced similar offerings that enable users to conduct general keyword searches within documents and ask questions about the content, Stravito said it sets itself apart by providing information from multiple relevant sources.
Instead of offering a single answer per document, Stravito’s answer engine delivers a comprehensive analysis. Moreover, the company claims that its answer engine is adept at handling complex datasets, even when they present conflicting information. For instance, the engine will identify and highlight the discrepancies if different sources provide varying numbers.
“Our solution is designed to help users avoid bias by providing context, and it doesn’t just share binary yes/no answers, but will share additional details — suggesting more questions for followup searches,” Thor Olof Philogène, CEO and founder at Stravito, told VentureBeat. “The answers are generated based on proprietary knowledge, not on unverified internet data, and customers’ sensitive information stays within the trusted enterprise cloud environment.”
VB Transform 2023 On-Demand
Did you miss a session from VB Transform 2023? Register to access the on-demand library for all of our featured sessions.
Enhancing traditional search through generative AI
The company said that its generative AI uses a client company’s internal knowledge base to generate answers. The tool will explicitly communicate if the platform lacks relevant knowledge to address a question. However, the user may still receive related context and existing information from the knowledge base.
>>Follow VentureBeat’s ongoing generative AI coverage<<
According to Philogène, although generative AI is exciting and promising, human intervention remains crucial. He emphasized that the summarized answers the AI generates should serve as a starting point, enabling users to quickly grasp key points, rather than being treated as the ultimate answer to a question.
“Direct links to source documents make it easy to fact-check as well as to dig deeper and uncover important nuances in the data,” Philogène told VentureBeat. “That’s why Stravito’s AI capabilities not only provide links to the sources but also explicitly highlight the exact paragraphs that were used to generate the answer.”
He said that in the realm of consumer insights and market research, the quality of answers depends heavily on the questions asked.
“While formulating the right question is a valuable skill, not all business stakeholders possess it, despite researchers investing significant time in developing this expertise,” explained Philogène. “Our AI’s recommended questions aim to solve both of these challenges: motivate users to continue exploration and learn more, but also help them ask better questions to get more concrete and helpful answers for further decision-making.”
Philogène said that his company strongly focuses on trustworthiness, data privacy, and security aligned with strict enterprise requirements, as the platform runs on enterprise cloud solutions with an SOC 2 Type II report.
“The agreements we have in place cover confidentiality, data ownership and intellectual property (so that no service provider will use customer content to train their model). As an ISO 27001-certified organization, we ensure high data privacy and security for our new generative AI features just like for the rest of our platform,” he said.
What’s next for Stravito?
The company said it would continue to develop generative AI capabilities based on customer feedback and needs, and plans to introduce several enhancements to the platform, including source control that allows users to exclude unwanted documents from generating answers. Furthermore, Stravito aims to integrate generative AI into existing workflows, such as Scrapbooks.
The company is exploring other areas in which to expand. These include providing concise one-page summaries for uploaded research, streamlining the desk research process through new production capabilities, and creating auto-generated summaries tailored to each user’s preferences.
“In the near future, AI-based communication will let us have meaningful talks and get personalized help from virtual assistants. At the same time, both new interfaces (from vendors) and new skills (from users) will be needed to make this cooperation seamless and effective,” said Philogène. ”At Stravito, we are very excited about the developments in generative AI and look forward to the groundbreaking developments we believe will happen in the market in the coming months and years.”
>>Don’t miss our special issue: Building the foundation for customer data quality.<<
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.