Active research and experiments with stated hypotheses and findings.
Multimodal ScamCheck — Screenshot & Image Scam Analysis (OCR + Vision + Semantic Retrieval)
Production multimodal scam-intelligence for ScamCheck: screenshot/image upload, lightweight OCR (Cloud Vision + Gemini fallback), deterministic fraud-signal detection, gated deep Gemini-vision analysis, and semantic comparison against known scam clusters via BigQuery VECTOR_SEARCH. Cost-gated, serverless, scale-to-zero.
ScamCheck Multimodal v3 — Production Evaluation Report
Large-scale evaluation of the ScamCheck multimodal scam-detection pipeline: a 1,000-sample synthetic corpus (en/hi/hinglish/mixed, 10 scam + 7 legit categories), precision/recall/F1, per-language and per-category breakdown, adversarial robustness, leaderboard analytics, caching/stress harnesses, cost model, scaling path, and known weaknesses.
Semantic Intelligence Platform — Retrieval, Enrichment, GEO & Scam Clustering APIs
The A Square Solutions semantic intelligence layer built on Vertex AI embeddings + BigQuery VECTOR_SEARCH: intelligent chunking, hybrid lexical+vector retrieval, snippets & confidence, semantic enrichment (topic/scam/trust/GEO), scam-pattern clustering, GEO/AI-search readiness scoring, and retrieval observability. Production, serverless, scale-to-zero, canonical 768-dim.
A Square Solutions — Full Blog & Page Content Audit (745 posts)
URL-level content audit of asquaresolution.com: 745 blog posts + 12 pages classified into keep/improve, merge, redirect, noindex, or support. Finds severe topical dilution (mostly off-topic AI/science/geopolitics news) and index bloat, with a consolidation plan, redirect map, internal-linking opportunities, and strategic manual-indexing priorities to rebuild topical authority around GEO/AI-SEO, AI automation, AI consulting, ScamCheck and TrustSeal.
A Square Solutions — Internal Linking Map, Topical Authority Clusters & GEO Structure
Semantic internal-linking architecture for asquaresolution.com: eight topical authority clusters (AI SEO, GEO, AI Automation, Technical SEO, Digital Marketing UK, AI Consulting, Scam Detection, Trust Verification), with pillar pages, service-to-blog and blog-to-service links, tool-to-service links, contextual anchor text, conversion pathways, and an entity-first GEO/AI-search structure. Implementation-ready for WordPress (Rank Math).
Live Pages — Optimization, Internal Linking & Schema/E-E-A-T Audit (verified)
Verified audit of asquaresolution.com live service pages and Tier-A assets: cannibalization/differentiation rules, a do-not-duplicate canonical list, title/meta/H/CTA optimization, an internal-linking additions matrix connecting GEO/AI-Automation/AI-Consulting/Technical-SEO/Entity-SEO + ScamCheck + TrustSeal, and a schema + E-E-A-T audit based on actual JSON-LD output (FAQPage/Service/Article/Person/Organization verified live). Includes sitemap gaps, author-identity standardization, indexing priority, and schema validation report.
5 entries
Complete visual evidence archive for the A Square Solutions Phase 2 WordPress ecosystem rollout (2026-05-20). Schema deployment, homepage sections, footer widget, About page, internal links, and LiteSpeed cache purge — every step captured with production screenshots and operational commentary.
Testing whether structured entity density and explicit answer formatting on asquaresolutions.com key pages increases AI citation frequency. Baseline established, implementation in progress.
Testing whether structured semantic HTML (dl/dt/dd elements with explicit field labels) increases AI crawler fact extraction accuracy compared to prose failure descriptions. The QuickFix component was designed as an operational hypothesis — this lab documents the reasoning, the implementation, and the observable indicators.
LiteSpeed's UCSS optimizer removes all CSS rules scoped to post-specific body classes (.postid-XXXX) because it crawls pages without those classes active. Any stylesheet-based per-post typography fix is silently neutralised.
Testing whether embedding an exact JSON schema + explicit format constraint in the prompt reduces malformed output frequency in Gemini 1.5-flash. Three prompt iterations tested during ScamCheck and TrustSeal build. Schema-in-prompt approach reduced parse failures from ~6% to <1% of calls.