Basylic (by Etaonis)

Basylic is a document-focused fraud detection platform developed by French startup ETAONIS (founded 2015).

Reviewed by 7wData

On this page

Publisher review

Basylic is a document-focused fraud detection platform developed by French startup ETAONIS (founded 2015). Rather than a generic risk-scoring engine, it specializes in detecting falsified or suspicious documents—government IDs, financial statements, tax records, invoices—across insurance claims, lending decisions, and public-sector benefit disbursement. The platform combines two layers: algorithmic detection (50+ proprietary models using deep learning, statistical anomaly detection, and computer vision) and human analyst review.

The hybrid approach is deliberately designed to minimize false positives in high-stakes use cases where a single missed fraud or incorrectly flagged claim can cause operational friction. Basylic targets regulated industries where document verification anchors the transaction: insurers checking policyholder identity and claim authenticity before payout; banks assessing applicant documentation during underwriting; real-estate firms conducting KYC on counterparties; government agencies protecting benefit and subsidy budgets. The product is particularly focused on French document types (RIB bank transfer records, national ID, tax notices), though the underlying architecture supports other formats.

The company emphasizes production maturity: ISO 27001 certification, RGPD compliance, and regular independent cybersecurity audits. It does not publish customer lists or case studies—typical for fraud tools serving sensitive institutions—which makes independent verification of effectiveness difficult. Weak points to watch: pricing is enterprise-only and not disclosed publicly, making it hard to assess total cost of ownership or compare against competitors without sales contact.

The geographic and document-type specificity means it is most defensible in French-speaking markets; organizations outside that footprint may find the value proposition narrower. There is no public evidence of G2, Gartner Peer Insights, or Reddit user reviews, limiting third-party validation of performance claims.

Get the AI & data signal, daily.

48k+ subscribers read this every morning. One email, both newsletters. Unsubscribe anytime.

How it works

  1. Generic and custom fraud controls

    Applies both universal anti-fraud rules and document-type-specific controls tuned by experts and clients.

  2. Human-in-the-loop analyst review

    Qualified fraud analysts pre-screen and interpret machine-generated alerts before delivery to end users.

  3. Multi-modal anomaly detection

    50+ proprietary deep learning, statistical, and computer vision models combined to surface inconsistencies and forgery indicators.

  4. Evidence-based report generation

    Automated reports with specific falsification indicators and confidence scores tied to underlying algorithmic signals.

  5. API integration and data enrichment

    Connects to external reference databases and internal business systems for context-aware verification.

  6. RGPD and ISO 27001 compliance

    Certified infrastructure with French hosting, regular independent security audits, and regulatory-ready data handling.

Strengths and trade-offs

Strengths

  • 10+ years of algorithm refinement with 50+ proprietary models specifically optimized for document-level fraud detection.
  • Human analyst layer designed to minimize false positives and ensure qualified, defensible output in regulated industries.
  • RGPD/ISO 27001 compliance and French data residency; purpose-built for insurance, banking, and public-sector buyers.

Trade-offs

  • Enterprise-only, quote-based pricing with no public tiers or calculator; ROI difficult to assess without direct vendor engagement.
  • No published customer roster or case studies; third-party validation limited to vendor-supplied claims and minimal G2/Gartner presence.
  • Strong geographic and document-type focus on French formats (RIB, national ID, tax notices); less defensible outside French-speaking markets or multi-language environments.

Pricing context

Enterprise SaaS model with sales-driven, quote-based pricing. No public pricing tiers, free tier, or self-service calculator available. Cost typically varies by document volume, fraud categories, and industry vertical. Typical for B2B fraud-detection tools targeting regulated sectors (insurance, banking, government) where contracts are negotiated per customer.

User reviews

No user reviews yet. Be the first to write one.

Sources

Reporting on this tool draws on these publicly available sources.

  1. www.basylic.fr — Official product website; confirms Basylic brand, parent company ETAONIS, core product positioning as document fraud detection platform.
  2. www.cbinsights.com — Company founding date (April 2015), headquarters (Levallois-Perret, France), business description, and investor/startup ecosystem presence.
  3. www.crunchbase.com — Verified company registration, founding year, employee count, business classification, and company profile data.
  4. pypi.org — Published Python API package; confirms technical integration capability, API-first architecture, and SDK availability for developers.
  5. www.linkedin.com — Company employment data, team credentials, and ongoing business activity verification.