Products
What we build
The governance suite serves the regulated-industry buyer. The CrawlQ family serves marketing and brand teams under its own brand.
Four products, three services and one knowledge graph, and what the board, auditor or regulator receives.
Source: The four products and three services as described on this site, 6 October 2026.
In detail
What each governance product does
The four products share one knowledge graph, so a finding in one is visible to the others.
TraceGov.ai
EU AI Act alignment scored across the five TRACE dimensions
Hosted platform, AWS Frankfurt (eu-central-1)
Transparency, Reasoning, Auditability, Compliance and Explainability, scored per AI system and backed by an audit trail an auditor can replay.
- Graph-based retrieval over the regulatory text
- Patent application EP26162901.8 (pending)
- Cryptographic audit trail, 7 year retention
GraQle
The intelligence engine
SDK on PyPI, v0.85.0, source-available, 13 model backends
The knowledge graph that connects the products. Multiple AI agents cross-check each other on governance questions. Source-available SDK on PyPI.
- Multi-agent cross-verification
- Zero LLM calls at retrieval
- Source-available: 2,009 tests
TAMR+
Trust-aware multi-signal document retrieval
TAMR+ Lite on PyPI (Apache 2.0); working paper on SSRN and Zenodo
The retrieval engine behind TraceGov.ai. Retrieval that understands the structure of regulations: articles, recitals and cross-references, not just keywords.
- Patent application EP26162901.8 (pending)
- TAMR+ Lite: standalone Python package on PyPI, Apache 2.0
- Working paper on SSRN and Zenodo
FrictionMelt
AI friction intelligence
Diagnostic framework: 95 friction points across 8 layers
Your AI pilots work. Enterprise rollout stalls. FrictionMelt diagnoses why: 95 friction points across 8 organisational layers with compound failure analysis.
- 95 friction points across 8 layers
- Compound failure trigger identification
- Prioritised by business impact
How GraQle refuses: seven hard gates are evaluated in order and a single FAIL stops the action (a design disclosure; the verdict schema is public from v0.84.1).
The gates
The seven hard gates; any single FAIL makes EXECUTE impossible, and each gate leaves a determinism record.
Source: github.com/quantamixsol/graqle · docs/dag/defensive-publication.md
How GraQle learns: the loop runs ANCHOR, ACTIVATE, GENERATE, VALIDATE and COMMIT through the project graph, and only a validated lesson returns to it.
The loop
The harness, the loop (ANCHOR, ACTIVATE, GENERATE, VALIDATE, COMMIT) and a real 146-node project graph, with a lesson validated and returned.
Source: github.com/quantamixsol/graqle (README; docs/dag/defensive-publication.md)
Own brand
The CrawlQ family
Marketing and brand tools under their own brand. Built by Quantamix, sold and supported at crawlq.ai and studio.crawlq.ai.
CrawlQ.ai
Content ERP
Philips cut 20 hours to 5 per product line and saved €500K in year one. CrawlQ turns scattered content into managed capital with 140+ psychographic factors and a full audit trail.
- 75% faster content cycles (Philips, verified)
- 2,800+ users including Fortune 500
- 95% brand compliance at scale
2,800+ users | 4.4 on Capterra
Studio CrawlQ.ai
Brand intelligence platform
Quantified Brand Trust Score, automatic deviation detection before publication, and AI-generated campaigns that follow the brand rules. Connected through the shared knowledge graph.
- Real-time Brand Trust Score
- Deviation detection before publication
- Shared knowledge graph with CrawlQ.ai
CopyNexus.io
AI copywriting at scale
Enterprise copy automation born from real production at Amazon Ring. Thousands of brand-consistent descriptions, social posts and listings in minutes.
- Built and tested in production at Amazon Ring
- Multi-channel: web, social, marketplace
- Connected to CrawlQ audience intelligence
Not yet released. Early access by email.