Cybersecurity startup Teleskope brings intelligent automation to data security and privacy compliance
Teleskope, a cybersecurity startup, is unveiling its data protection platform, engineered to automate data security, privacy, and compliance at scale. Backed by its $2.2 million pre-seed funding led by Lerer Hippeau, Teleskope tackles false positives associated with traditional Data Security Posture Management (DSPM) to drive scalable security without increasing manual and operational burdens.
Teleskope’s platform works by employing AI and large language models (LLMs) to eradicate constant, manual assessments that have long been integral to data security. By automating this aspect of security, Teleskope replaces point-in-time spreadsheets and ad hoc scripts with advanced, real-time security and privacy posture.
With its founders empowered by this detrimental inefficiency plaguing many organizations, Teleskope offers advanced contextual analysis that goes beyond traditional data classification, according to the vendor.
Teleskope uses an advanced classification engine, powered by a LLM and rules engine, to identify whether specific data—such as a customer’s home address, a business address, or public landmark—is considered Personally Identifiable Information (PII). By monitoring cloud data stores and third-party vendors, Teleskope offers its users a thorough inventory of assets, including those that are seemingly hidden, and identifies its security and compliance risks.
“What sets Teleskope apart as a data security and protection solution is its 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 Nancy Wang, director of engineering for data protection and security at AWS. “Teleskope enables engineers to take on a much bigger role when it comes to protecting sensitive business and customer data."
According to the company, Teleskope’s LLMs are consistently updated and optimized for speed and cost, further driving its ability to identify data subjects, which can include a customer or employee, as well as information about customer types.
While traditional DSPMs tend to break as proprietary data begins to scale and lack contextual understanding, Teleskope ensures accurate insights that scale as an enterprise does. This is done while integrating seamlessly with existing workloads and developer pipelines, enabling organizations to avoid data breaches and privacy incidents with ease.
“A lot of people have been jaded by data classification tools; they turn them on, and then they turn them off,” said Elizabeth Nammour, co-founder and CEO of Teleskope. “I really want them to think that ‘Okay, it is truly possible to have a good data classification tool that can actually reduce my workload instead of adding on to it.’ We're really excited about reducing alert fatigue and providing actually good results to our users.”
The platform only gets increasingly accurate with its usage; the Teleskope AI model is trained to learn each individual customer implementation to drive its overall precision. Teleskope automatically applies compliance enforcement at the source, accompanied also by the capability for developers to implement custom security and privacy protocols via open APIs, according to the vendor.
“Teleskope actually works in doing what it says it does, which is helping you reduce your security and privacy vulnerabilities and get an understanding of your data,” said Nammour. “They can [also] automate the manual work; a lot of us engineers out there are stuck doing a lot of things like manual reviews, manual assessments. [With Teleskope,] they'll be able to automate the awful work that they have to do day-in, day-out that's manual.”
Teleskope’s data protection platform supports both structured and unstructured data stores across various cloud platforms, including AWS, GCP, and Snowflake. Additionally, the platform supports third-party SaaS, able to identify over 100 data types.
“We can deploy our large language model within their [a customer’s] cloud environment, so the data never leaves their cloud,” explained Nammour. “Outside of that, we run Teleskope as a single-tenant solution. So, for each of our customers, we spin up a new large language model for each of them separately, so that the data never gets commingled or shared. We process it, but then we never store a copy of the data. We just process it in-memory.”
In the future, Teleskope will delve deeper into the identification details it is capable of, as well as its ability to automate and remediate issues on its own.
“We really want to help you go into the nitty gritty and help you understand, [for example,] if you have five different types of customers, such as patients and doctors, we want to help you differentiate that within your data,” said Nammour. “Right now, our customers can automate their own [remediation,] but eventually, we'll want to be able to do that automation ourselves in a way that actually works—by plugging into their infrastructures and code.”
To learn more about Teleskope’s platform, please visit https://www.teleskope.ai/.