Teleskope raised $2.2 million in pre-seed funding led by venture capital fund Lerer Hippeau. The company’s automated data protection platform assists organizations in complying with regulations such as GDPR and CCPA and reduces the manual burden on security, data, and engineering teams.
Unlike traditional data security posture management (DSPM) approaches that often result in alert fatigue due to valuable time wasted on false positives, Teleskope uses large language models (LLM) to provide actionable insights with greater accuracy. The company’s data protection software integrates with existing workflows and developer pipelines, giving engineers the ability to automate security at the source.
By minimizing false positives and providing context around the data, Teleskope empowers organizations to focus on genuine vulnerabilities – streamlining their security and privacy operations while significantly reducing risk – particularly important at a time when data breaches and privacy incidents are increasing, along with fines for improper handling of data.
According to a research report published by Fastly in partnership with Enterprise Strategy Group, 75 percent of organizations spend as much time, and sometimes more time, on false positives compared to actual attacks. These false alerts cause the same amount of downtime as real attacks. Tooling fatigue and false positives are impeding ability to maintain adequate security across new and existing application architectures.”
Founded by two former Airbnb security engineers, Elizabeth Nammour and Julie Trias, Teleskope brings the combination of expertise and fresh perspective to the issue of data security. Prior to starting Teleskope, the co-founders wrestled with the problem of constant manual assessments, and reviews that become obsolete as soon as they’re completed. They recognized the need to replace point-in-time spreadsheets and ad hoc scripts with automation that provides a real-time and always up-to-date data security and privacy posture.
“With the increasing use of generative AI causing even more data leaks, privacy is becoming even more of a concern. And we know from our work at Airbnb that data classification products don’t work the way they should – too many false-positives, a lack of contextual understanding, and an inability to scale and integrate causes alert fatigue and puts critical data at risk. “Now more than ever, privacy and security concerns are escalating due to the widespread use of generative AI,” said Elizabeth Nammour, co-founder and CEO of Teleskope. “Teleskope helps customers automate security and privacy saving them valuable time, while simultaneously reducing risk. From detection to remediation, Teleskope can be easily integrated in developer pipelines to help prevent and resolve data security vulnerabilities.”
Teleskope’s advanced contextual analysis goes beyond traditional data classification to discern whether data – such as an address – is actually considered personally identifiable information (PII), when associated with a customer’s home address, or when that address is associated with a public landmark, or business address, and does not have PII implications. Teleskope uses state-of-the-art large language models, which are continuously fine-tuned and optimized for speed and cost, to identify the data subject associated with the data, such as a customer, employee, or differentiate between customer types, such as doctors and patients.
“The cyber field needs more diversity of thought, as well as leaders building with hands-on experience as engineers and operators,” said Graham Brown, managing partner at Lerer Hippeau. “Teleskope’s founders’ extensive experience as security-focused software engineers, including for the military, brings fresh perspective, as they’ve lived through the pain points they’re solving. Plus, their developer-centric approach and strategic use of AI for best-in-class contextual classification position them to build the industry-leading platform to protect and secure customer information.”
The company has assembled a team of industry experts as advisors, which includes Doug Dooley, chief operating officer of the application security company Data Theorem, Abby Kearns, former chief technology officer of Puppet, Sathia Narayanan Mahadevan, head of privacy and assurance at Reddit and Nancy Wang, director of engineering for data protection and security at Amazon Web Services.
“What sets Teleskope apart as a data security and protection solution is their unique ability to detect and remediate security vulnerabilities before they enter production, by enforcing protection policies at every stage of the software development lifecycle (SDLC),” said Wang. “Teleskope enables engineers to take on a much bigger role when it comes to protecting sensitive business and customer data.”
“Teleskope has been a game-changer for us in streamlining our compliance and security processes,” said Nick Cobb, vice president of engineering and product at Kyte, a service for cars delivered on demand. “We’ve found Teleskope to be incredibly accurate in classifying data, and it has helped us uncover customer PII and sensitive information that was previously overlooked through manual labeling. By integrating Teleskope’s findings into our existing workflows through APIs, our engineers were able to automate compliance tasks, saving them valuable hours of operational overhead, and allowing them to focus on their core responsibilities. As an early-stage company with a complex on-the-ground business, this is invaluable to us in the time of limited capital and resources.”
Teleskope can, within minutes, monitor cloud data stores and third-party vendors, providing a comprehensive inventory of assets, including hidden ones, and identify their security and compliance risks. Its advanced classification engine, powered by a large language model and rules engine, is able to adapt to each unique data store – identifying sensitive data and providing actionable context on who the data is about. The Teleskope AI model is trained to learn each unique customer implementation to further increase its accuracy.
