Fundamentia: IDP and Anonymization for Regulated AI

Page contributed by Francisco Morales Arcia.

Spanish deep-tech vendor pairing SimplÿData, a hybrid LLM and pattern-based intelligent document processing platform, with PrivacÿShield semantic anonymization, aimed at regulated banking, insurance, and legal workflows in the EU.

Overview

Fundamentia (legally Fundamentia Business Consulting SL) is a Spanish deep-tech company headquartered in Elche, Alicante, that builds intelligent document processing for regulated industries. Founded in 2013 and operating a team of around fifteen people, the company positions itself not as a pure software vendor but as an operator that designs, deploys, and runs end-to-end document automations tuned to each customer's existing systems. Its stated markets are banking and finance, insurance, legal, real estate, healthcare, and education.

The company's core product, SimplÿData, traces back to an EU Horizon 2020 grant (project 867653, "Unlocking unstructured data towards intelligent work automation for the banking industry"), and the platform has since been recognized with an EU Seal of Excellence. According to COO Francisco Morales Arcia, SimplÿData processes roughly 1.2 million pages per month in production at several of Spain's largest banks and additional mid-market financial institutions, with customer-reported reductions of 60–80% in document-processing time and cost. Fundamentia's own website cites a higher figure of 5M+ pages per month; the ~1.2M figure is the one its COO states publicly and is used here.

Spain is Fundamentia's core market, with expansion into further European markets described as the next step and Spanish-speaking Latin America as a mid-term target. Customer names are held under contractual confidentiality; reference calls are offered only within concrete evaluations, so the deployment claims above rest on vendor statements rather than named, verifiable references.

How Fundamentia processes documents

SimplÿData identifies, classifies, and extracts data from documents such as deeds, contracts, emails, and reports using natural language processing and supervised machine learning. The vendor reports document identification and classification accuracy exceeding 99% and data extraction accuracy exceeding 95% (97%+ on its current site). These are vendor-published figures; no independent benchmark of SimplÿData accuracy is currently available.

Architecturally, the platform is described as a hybrid that pairs large language models for language understanding and flexibility with a pattern-based layer for precision, determinism, and cost control. Fundamentia presents this determinism as a requirement for regulated customers who cannot rely on probabilistic output alone, and supports on-premise and local-inference deployment so sensitive data can remain inside the customer's perimeter. The platform is OCR-agnostic, sitting on top of whichever OCR engine a customer already uses rather than mandating its own.

SimplÿData is designed to complement rather than replace existing automation. It exposes REST APIs alongside its own connectors for legacy systems, document management systems, BPM platforms, and RPA tools, so extracted and validated data can feed downstream workflows with traceability and audit logging.

PrivacÿShield: semantic anonymization

PrivacÿShield is Fundamentia's second product, available standalone as SaaS and able to run on any document source independently of SimplÿData; used together, the two cover end-to-end document workflows. It performs anonymization of personal, financial, and sensitive data, with a vendor-claimed 99%+ accuracy.

Its differentiator is what Fundamentia calls semantic anonymization: rather than only removing explicit identifiers (names, national IDs, IBANs, card numbers), it tracks a person across a document and masks indirect mentions such as a job title, a city, a date, or a reference that would otherwise allow re-identification from context. The company frames this as true GDPR anonymization as opposed to mere pseudonymization, and notes that a properly anonymized document can retain analytical utility for studies, analytics, or model training without re-identifying real individuals. As with the accuracy figures, this positioning is vendor-stated and has not been independently tested here.

Use cases

Banking and mortgage processing

Fundamentia's strongest claimed proof point is mortgage processing automation for Spanish banks and outsourcing providers, where SimplÿData processes documents from effectively every major Spanish bank. Deeds, contracts, and mortgage files are classified and extracted to feed back-office and origination workflows.

Insurance, legal, and back-office review

Beyond banking, the platform targets document-intensive onboarding, back-office, and regulatory review across insurance, legal, real estate, healthcare, and education. These are processes where the vendor emphasizes that errors carry operational, financial, or regulatory impact.

Privacy and data reuse

Through PrivacÿShield, Fundamentia targets teams that need to share or analyze sensitive documents for analytics, biomedical studies, or AI training while reducing re-identification risk under GDPR, positioning anonymization as a way to give documents a "second life" rather than only a protective control. Teams evaluating GDPR-focused, EU-hosted alternatives may also consider Tiny IDP, a Spain-based extraction API with zero data retention by default.

Technical specifications

Resources

• Website
• PrivacÿShield
• Horizon 2020 project (CORDIS)
• LinkedIn

Company information

Calle Santiago Ramón y Cajal 41
03203 Elche (Alicante), Spain
Fundamentia Business Consulting SL (CIF B86644598)
www.fundamentia.com
LinkedIn