Junyul is an enterprise platform designed to provide operational evidence management for organizations running AI systems, with a particular focus on compliance with regulations such as the Korean AI Basic Law and the EU AI Act. The platform collects and links operational evidence required for regulatory response, operating under a principle that original data is not transmitted to Junyul servers by default. Instead, its SDKs generate local hashes and only send evidence fingerprints and minimal metadata to the server, supporting privacy and data minimization.
The platform is built to structure and connect asset inventories, impact assessments, incident records, and audit reports, enabling security, legal, platform, and executive teams to make decisions based on the same underlying evidence. Junyul provides a unified evidence graph that covers the entire AI operational lifecycle, including asset registration, impact classification, runtime evidence collection, regulatory change tracking, reporting, and auditing. It supports the detection and investigation of incidents such as prompt injection, tool misuse, memory poisoning, and suspected data leaks, reconstructing event timelines and forensic evidence for analysis and reporting.
Junyul offers SDKs in Python, TypeScript, and Go, distributed via PyPI, npm, and Go module, which instrument AI inference, tool calls, searches, and automated decisions as events. The platform integrates with enterprise workflows through API keys, webhooks, Slack, SIEM, and supports OpenTelemetry. Its Incident Command Dashboard allows collaborative investigation and reporting for security, SRE, and legal teams.
The system is intended for organizations needing to demonstrate compliance with multiple frameworks and regulations, including the Korean AI Basic Law, EU AI Act, NIST AI RMF, OWASP Agentic Top 10, and privacy laws. It does not provide legal advice but structures evidence to support legal and compliance assessments. Pricing and contract terms are determined based on organizational adoption scope, and operational use is controlled via workspace API keys and paid contracts.
In the LLM eval & observability space, Junyul takes a focused approach. It focuses on managing and documenting AI compliance, risk, and regulatory evidence for enterprises. Junyul is a B2B product aimed at enterprise compliance and risk teams. Pricing is enterprise-only. Junyul is available on the web and the command line.
Behind Junyul is Junyul, based in South Korea, and the product first shipped in 2026. Among its 5 catalogued features are AI asset inventory, impact assessment, and incident investigation. The interface is available in English and Korean.
Latest indexed changes and source events
Other apps tracked under the same category.
