TrustAI: A Multi-Source Artificial Intelligence Verification Platform with Authority-Weighted Trust Scoring, Retrieval-First Caching, and Domain-Adaptive API Routing
Fuente:
Zenodo
Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | English |
| Published: |
Zenodo
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866901923821518848 |
|---|---|
| author | Egbedayo, Oyelokiki George |
| author_facet | Egbedayo, Oyelokiki George |
| contents | <p>TrustAI is a production-deployed artificial intelligence verification platform developed by Rainfame Technologies Limited. The platform addresses the inability of end users to assess the factual accuracy of AI-generated content by implementing a multi-source verification pipeline that submits natural language queries simultaneously to a plurality of AI inference engines and authoritative external knowledge APIs, computes an authority-weighted trust score from the degree of inter-source consensus, and delivers verified responses via a server-sent event streaming protocol.</p> <p>Four original technical innovations are documented and protected by UK patent applications filed 14 May 2026. First, a multi-source AI verification pipeline with dynamic authority-weighted trust scoring wherein government primary sources are assigned weights of 0.90 and established reference sources 0.55 (UK Patent GB2611334.0). Second, a retrieval-first response caching architecture using 1536-dimensional dense vector embeddings and cosine similarity search with multi-gate quality filtering enforcing distance, trust score, and temporal freshness constraints (UK Patent GB2611337.3). Third, a domain-jurisdiction adaptive API routing system that selects institutionally relevant authoritative sources based on inferred query semantic domain and legal jurisdiction using a priority-ordered configuration table (UK Patent GB2611338.1). Fourth, a two-tier asynchronous streaming architecture delivering initial response tokens within approximately two seconds via a fast-path model while executing the full verification pipeline concurrently in the background (UK Patent GB2611339.9). A fifth innovation covering a privacy-preserving multi-provider architecture with user-scoped vector isolation, substitution-based query anonymisation, and plan-aware data governance is covered by UK Patent GB2611340.7.</p> <p>The platform is deployed in production at rainfame.com serving users across thirteen professional domain modes including legal, compliance, medical, accounting, finance, government, research, real estate, insurance, banking, human resources, supply chain, and IT security. Authoritative source integrations include legislation.gov.uk, EUR-Lex, Laws.Africa, Wikipedia, Wikidata, and Companies House UK.</p> <p>All five UK patent applications filed 14 May 2026 at the Intellectual Property Office, Newport, Wales under the Patents Act 1977. Applicant: Oyelokiki George Egbedayo.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_20204378 |
| institution | Zenodo |
| language | eng |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | TrustAI: A Multi-Source Artificial Intelligence Verification Platform with Authority-Weighted Trust Scoring, Retrieval-First Caching, and Domain-Adaptive API Routing Egbedayo, Oyelokiki George Artificial Intelligence Verification Trust Scoring Hallucination Detection Multi-source AI Authority Weighting Local Technology Compliance Technology Vector Memory Domain Adaptive Routing Professional AI Deployment Regtech Natural Language Processing <p>TrustAI is a production-deployed artificial intelligence verification platform developed by Rainfame Technologies Limited. The platform addresses the inability of end users to assess the factual accuracy of AI-generated content by implementing a multi-source verification pipeline that submits natural language queries simultaneously to a plurality of AI inference engines and authoritative external knowledge APIs, computes an authority-weighted trust score from the degree of inter-source consensus, and delivers verified responses via a server-sent event streaming protocol.</p> <p>Four original technical innovations are documented and protected by UK patent applications filed 14 May 2026. First, a multi-source AI verification pipeline with dynamic authority-weighted trust scoring wherein government primary sources are assigned weights of 0.90 and established reference sources 0.55 (UK Patent GB2611334.0). Second, a retrieval-first response caching architecture using 1536-dimensional dense vector embeddings and cosine similarity search with multi-gate quality filtering enforcing distance, trust score, and temporal freshness constraints (UK Patent GB2611337.3). Third, a domain-jurisdiction adaptive API routing system that selects institutionally relevant authoritative sources based on inferred query semantic domain and legal jurisdiction using a priority-ordered configuration table (UK Patent GB2611338.1). Fourth, a two-tier asynchronous streaming architecture delivering initial response tokens within approximately two seconds via a fast-path model while executing the full verification pipeline concurrently in the background (UK Patent GB2611339.9). A fifth innovation covering a privacy-preserving multi-provider architecture with user-scoped vector isolation, substitution-based query anonymisation, and plan-aware data governance is covered by UK Patent GB2611340.7.</p> <p>The platform is deployed in production at rainfame.com serving users across thirteen professional domain modes including legal, compliance, medical, accounting, finance, government, research, real estate, insurance, banking, human resources, supply chain, and IT security. Authoritative source integrations include legislation.gov.uk, EUR-Lex, Laws.Africa, Wikipedia, Wikidata, and Companies House UK.</p> <p>All five UK patent applications filed 14 May 2026 at the Intellectual Property Office, Newport, Wales under the Patents Act 1977. Applicant: Oyelokiki George Egbedayo.</p> |
| title | TrustAI: A Multi-Source Artificial Intelligence Verification Platform with Authority-Weighted Trust Scoring, Retrieval-First Caching, and Domain-Adaptive API Routing |
| topic | Artificial Intelligence Verification Trust Scoring Hallucination Detection Multi-source AI Authority Weighting Local Technology Compliance Technology Vector Memory Domain Adaptive Routing Professional AI Deployment Regtech Natural Language Processing |
| url | https://doi.org/10.5281/zenodo.20204378 |