Rising Star Nymiz

Short Summary

Nymiz is a Series A-funded privacy technology startup founded in 2020 and headquartered in Bilbao, Spain, focused on unstructured data de-identification for documents, scans, images, and selected database fields. Its platform detects personal data and then anonymizes or pseudonymizes it so organizations can safely share content with analytics systems, AI workflows, or third parties. The strongest fit is in public administrations, courts, law firms, and utilities—especially where large volumes of scanned material and European languages (including Latin American variants) create operational and compliance pressure.

The product’s core innovations are a patented normalization layer that converts many file types into a single internal representation, and proprietary small language models (SLMs) that run on standard CPUs, enabling on-premises operation and keeping processing inside the customer boundary. Detection is augmented with clustering, question-answering models, and computer vision for items like QR codes, barcodes, logos, and handwritten signatures. Treatment supports irreversible anonymization or reversible pseudonymization, with a customer-controlled encrypted re-identification dictionary; it writes results back into the original file so it remains usable in its native application, including proper PDF redaction that removes underlying content.

Nymiz positions itself as the remediation step that complements discovery, governance, DSPM, DLP, and AI governance tools rather than replacing them. Integrations (notably a live BigID plugin), REST APIs, and an MCP server enable pipeline-based de-identification, including “privacy gateway” use for AI agents. Challenges include limited native coverage beyond personal data, reliance on partner platforms and still-developing repository connectors (only Google Drive is live), and a customer footprint concentrated in Southern Europe and Latin America. The roadmap emphasizes broader connectors, an SDK, and scaling go-to-market reach.

Interesting Facts

• Nymiz was founded in 2020 and is headquartered in Bilbao, Spain.
• Nymiz has raised approximately $6 million to date, including a €2 million round led by TIN Capital in September 2025.
• The company targets customers primarily in Europe and Latin America.
• The platform detects personal data in documents, scanned pages, images, and databases and then anonymizes or pseudonymizes it.
• Nymiz runs proprietary small language models on standard CPUs without requiring GPUs.
• Google Drive is the only live repository connector.
• SharePoint, OneDrive, Amazon S3, Microsoft Purview, and iManage connectors are in development.
• Deployment options include on-premises, single-tenant, and multi-tenant SaaS, certified to ISO/IEC 27001 and ENS High.
• A BigID plugin integration is live.
• A date of birth, a postcode, and a gender are often enough to single out one person.

Notable Quotes

• “Finding sensitive data is largely a solved problem; treating it automatically, at the pace of AI pipelines, is not.”
• “A date of birth, a postcode, and a gender are often enough to single out one person.”
• “Removing names rarely makes a document anonymous.”
• “Nymiz argues that enterprises are moving from a need-to-know posture to a need-to-share one, because data held back today will eventually be needed by an analytics workflow, a model, or a third party.”
• “The score supports review, but it neither guarantees anonymity nor removes the need for human review and broader data security controls.”
• “De-identification must run as a step inside those pipelines, triggered by whichever system classifies the data.”
• “The treated file must still open in the application that produced it.”
• “The discipline borders on data discovery and classification, data masking, and synthetic data generation.”
• “The reversible workflow has also gained legal relevance.”
• “From our perspective, unstructured data de-identification is likely to be absorbed into data security and AI governance architectures as an embedded service.”