🔵 NEO Update: Software Intelligence That Fills the Gaps

Product Update • Q4 2025

NEO Intelligence Platform Enhancements

Introducing major enhancements to NEO product recognition and data enrichment: 90% recognition rate through multi-layer intelligence, smart version inference that eliminates blind spots when vendor data is incomplete, 70% smaller footprint through lifecycle grouping, and automatic AI capability detection, all grounded in public, verifiable sources.

90%recognition rate
70% smallercatalog footprint
100%source tracking
AI-awarecapability detection

What’s New

🔵 NEO AI Capability Detection NEW

NEO now automatically identifies products with AI/ML capabilities and tracks vendor data training policies:

Example: Microsoft 365 Copilot Detection
Detected Attributes:
  • has_ai_capabilities: true — Product includes generative AI features
  • uses_user_data_for_training: false — Microsoft’s public commitment: no customer data for training
  • Source: microsoft.com/privacy/copilot
  • Confidence: 95%
Example: Adobe Creative Cloud AI Features
Detected Attributes:
  • has_ai_capabilities: true — Firefly, neural filters, content-aware fill
  • uses_user_data_for_training: true — Opt-out required for content credentials
  • Source: adobe.com/ai-guidelines
  • Confidence: 92%

This helps organizations manage AI governance, data privacy compliance, and licensing for AI-enhanced products.

Platform Improvements

🎯 90% Recognition Based on Multi-Layered AI Recognition

Products are recognized through multiple intelligent matching techniques working in concert, from exact synonym matches to AI-powered semantic analysis, ensuring we identify even poorly-formatted, abbreviated, or vendor-specific product names.

This multi-layered approach achieves 90% recognition rates compared to typical 70-80% with traditional exact-match systems.

🔮 Smart Version Inference: Filling Vendor Data Gaps

The breakthrough: when vendors don’t publish lifecycle data for every patch version, we intelligently infer it from known lifecycle boundaries.

Example: Node.js 6.17.1 (Missing Lifecycle Data)
Scenario: Legacy Node.js version still in production, but vendor removed v6 documentation from official pages.
Vendor’s EOL Page:
Node.js 6 → No data found
⚠ Version 6 not listed
Node.js 7 → EOL Jun 2017
Next version available
NEO Enrichment:
Node.js 6.17.1 → EOL Jun 2017 Inferred (78% confidence)
✓ Inferred from v7 (higher version)
Example: Ubuntu 22.04.3
Scenario: Point releases like 22.04.3 inherit LTS support windows from the major release.
Before:
Ubuntu 22.04.3 Unknown EOL
⚠ Security blind spot
After:
Ubuntu 22.04.3 → EOL Apr 2027 Matched (100% confidence)
✓ From 22.04 LTS lifecycle
Example: Microsoft Visual C++ Redistributable 14.38.33130
Scenario: Critical runtime library deployed everywhere, but Microsoft only publishes lifecycle for major versions.
Vendor Documentation:
VC++ 2015-2022 → Lifecycle tied to Visual Studio 2022
No patch-level EOL data
NEO Enrichment:
VC++ 14.38.33130 → EOL Apr 2027 Inferred (88% confidence)
✓ From Visual Studio 2022 lifecycle
Result: Zero security blind spots. Every version gets a lifecycle status, even patch releases vendors don’t explicitly document.

📦 70% Smaller Catalog Footprint

By grouping versions based on lifecycle boundaries rather than tracking every individual patch, NEO dramatically reduces catalog size while maintaining complete matching accuracy. Thousands of version strings collapse into manageable lifecycle groups, making queries faster and simpler.

Example: 2,000+ product variants consolidate to 347 unique products. Catalog updates complete in minutes instead of hours.

Grounded in Public Sources

Every lifecycle date, every inference, every AI capability flag comes from verifiable public sources:

🥇 Primary Sources
• Official vendor lifecycle pages
• Product documentation
• Security advisories
• Support policy pages
🥈 Secondary Sources
• NVD (National Vulnerability DB)
• endoflife.date
• Official GitHub releases
• SPDX registries
🚫 Never Used
• Random blogs
• Community forums
• Speculation/rumors
• Paywalled databases
Source Tracking Example
// Every catalog entry includes:
product: “PostgreSQL”
version: “15”
eol_date: “2027-11-11”
confidence: 0.98
sources: [
“postgresql.org/support/versioning/”,
“endoflife.date/postgresql”
]
status: “matched” // or “inferred”

Why This Matters

For Security Teams

  • Zero blind spots: Version inference ensures every product in your software inventory has lifecycle data, even obscure patches
  • Risk prioritization: Confidence scores help focus on high-certainty threats first
  • Audit trail: Every EOL date links back to official vendor sources

For IT Asset Management

  • Reduced complexity: 70% smaller footprint means faster queries and simpler reporting across all devices and software
  • Better coverage: 90% recognition rate vs typical 60-70% with exact-match-only systems
  • AI transparency: Know which products have AI features and data policies

Real-World Results

<1 day
Process 100K+ products from raw software inventory to complete catalog
vs 6-12 months manual research with missing gaps at version level
Daily updates
Automated catalog refreshes keep lifecycle data current across devices and software inventory
Stay ahead of EOL dates and security changes
1,200+
Authoritative public sources tracked for lifecycle validation
100% audit-ready provenance

Getting Started

The enhanced NEO engine is available now. Existing catalogs will automatically benefit from version inference and AI capability detection on the next scheduled refresh.

Posted in

Alex Cojocaru

Alex has been active in the software world since he started his career as an Analyst in 2011. He had various roles in software asset management, data analytics, and software development. He walked in the shoes of an analyst, auditor, advisor, and software engineer, being involved in building SAM tools, amongst other data-focused projects. In 2020, Alex co-founded Licenseware and is currently leading the company as CEO.