Uptycs, which provides a unified cloud native application protection platform (CNAPP) and extended detection and response (XDR), has launched a “no slop” AI security analyst, Juno, designed to bolster verifiable protection.
“Most of the time, uncertainty rules in the world of cybersecurity,” said Ganesh Pai, CEO of Uptycs. “Juno’s evidence-based approach uses AI to replace opaque ‘black box’ answers with transparent, verifiable reasoning grounded in real telemetry, so security teams can trust what they’re seeing and act with confidence. By showing not just what to do but why, Juno helps teams move faster without taking unnecessary or disruptive risks.”
Uptycs officials argued that most security AI operates as a “black box,” offering probability-based guesses without proof. In contrast, Juno utilizes a “glass box” approach. In generating answers, it executes deterministic SQL queries against a purpose-built unified multi-cloud ontology of more than 3,000 tables and 150,000 columns of security telemetry.
“This architecture allows Juno to take surgical sips of data, retrieving precise, raw evidence to prove its findings, rather than ingesting a firehose of noise that leads to hallucinations,” officials said.
Uptycs believes there is a gap in the CNAPP market because some solutions rely on “federated” data architectures with separate databases for cloud, endpoint and identity glued together via APIs, which makes it difficult to execute verifiable, cross-platform AI with precision.
“Because Uptycs normalizes all CNAPP telemetry into a single language, Juno can answer open-ended questions across complex infrastructure and verify citations against industry bodies like CVE and vendor whitepapers,” officials said.
Thus far, Juno has been deployed by major automotive manufacturers, banks and enterprises, the company said.
