Reko, a company that uses AI to map company data exchanges, today announced it has raised $30 million in a Series A round led by Insight Partners, involving Zeev Ventures, BoldStart, Angular Ventures, Jibe Ventures, CrewCapital and Cyber Club London. CEO Ofer Klein said the proceeds will be used to develop the product and support the company’s go-to-market efforts.
Klein co-founded Reco with Gal Nakash and Tal Shapira in 2020. Reco is Klein’s second venture after Kwik, an IoT platform for “connected customer experience.” Nakash led research at the Israeli Prime Minister’s Office before joining Reco, and Shapira, who also worked in the Prime Minister’s Office, was head of algorithms at heads-up display startup Guardian Optical Technologies.
“The distributed workforce is getting bigger. The number of collaboration tools used in a single organization is increasing every day. And each of them presents a new security threat,” Klein told TechCrunch in an email interview. “Information sharing and collaboration across applications is critical to and of unlimited value to businesses, but without the security context and intelligence of what business data and interactions occur in these exchange environments, we see organizations increase their risk.”
Reco monitors communication within and outside the organization to flag potential issues when they occur. For example, if a contract with a third-party vendor has expired but the vendor still has access to confidential documents, or if a file is sent to the wrong account manager, Reco can detect it and flag it for correction, Klein said.
There are many compliance and data management software available (see: Checks, DataGuard, ketch as well as DataGrail), but Klein argues that Ryoko’s “contextual” approach sets him apart. The platform creates a “business collaboration map” that purportedly shows where the incident happened, which parties were involved, and details of remedial actions.
According to Klein, Reco uses an artificial intelligence engine to learn context through metadata in audit logs generated by collaboration, storage, and email tools (such as Slack, Jira, Box, OneDrive, Outlook, etc.). The mechanism does not require manual policy configuration to create a relationship model between organization employees, teams, and external vendors, Klein said, freeing up security teams to focus on more pressing work.
“Our closest competitors are legacy data protection tools that are still focused on content, not context. They require a high level of manual entry and maintenance, create noise and organizational friction, and have no understanding of the context of incidents leading to inaccurate alerts,” Klein said. “[With Reco,] security teams can share incidents and their context with the appropriate team members to get additional feedback to fix the incident, or even give a team member the opportunity to take ownership of the incident and take corrective action on their own.”
Of course AI does mistakes. Unforeseen biases can affect the system’s predictions, causing mistakes. The scale of the problem is evidenced by a 2021 survey conducted by Fastly and ESG. found that nearly half of all cybersecurity alerts from software are false positives.
For what it’s worth, Klein stands by his product, stating that Reco “results in the fewest false positives in detecting events and incidents that no other tools can detect.”
“The pandemic has pushed companies to make extensive use of collaboration tools to support their distributed workforce. Companies have started using tools like Slack, Teams and Google Drive with their customers and vendors to speed up business processes,” Klein said. “But there is no visibility and limited control over these channels… Reco is well positioned to lead the next generation of security solutions that are focused on growing businesses while saving significant money and effort.”
Klein declined to list the names of Reco’s 35-employee clients or disclose earnings, but said “many” companies in finance, manufacturing, technology and insurance, including Fortune 500 companies, use his product.
Credit: techcrunch.com /