Squin is a technical consultancy and research publication specializing in semantic SEO, structured data, and entity optimization.
In short, we help diagnose and fix how B2B SaaS brands are represented by AI systems – when ChatGPT, Perplexity, Gemini, and AI Overviews miscategorize, omit, or hallucinate the products buyers are researching.
The problem this site exists to solve
In 2026, B2B software buyers don’t start their vendor research in Google. They ask ChatGPT, Perplexity, and Gemini “what’s the best [category] platform for [use case]” – and the answer they get determines which vendors land on the shortlist before a marketing team has a chance to influence it.
The signals these systems use to identify, categorize, and represent brands are entity-based: structured data (JSON-LD), knowledge graph presence (Google KG, Wikidata), category coherence across review sites and authoritative sources, and consistent semantic positioning across the open web.
The truth is, most mid-market SaaS companies have broken or absent implementations across all four layers. The result is three commercial harms:
- Brands get cited inconsistently in vendor-research queries, or not at all.
- Categories get collapsed – a “customer success platform” described as a “helpdesk tool,” wiping out positioning that took years to build.
- And AI systems hallucinate attributes, features, or pricing when no authoritative entity grounding exists.
Squin diagnoses and fixes this.
Who runs Squin
Squin was founded by Ričerdas Dapkevičius and operates as a boutique consultancy: a small senior team focused on entity optimization for B2B SaaS.
The work sits between technical implementation and brand positioning – JSON-LD architecture, knowledge graph engineering, and entity coherence across the open web. Squin exists because the same pattern keeps surfacing in mid-market B2B SaaS: marketing teams treating AI search visibility as a content problem when it’s an infrastructure problem. The fix isn’t a better blog post. It’s a coherent entity graph the CMO didn’t know they needed.
A note from the founder:
I started Squin after watching the same diagnostic conversation play out across a dozen SaaS marketing teams – and watching their existing SEO agencies fail to even name what was breaking. The vocabulary gap is the opportunity. Squin is built to close it. — Ričerdas Dapkevičius
How Squin started
Squin.org was originally home to the SQUIN Project – a Traversal-Based Query Execution system for the Semantic Web, developed in academic research and cited across peer-reviewed computer science papers. The original SQUIN let researchers run SPARQL queries across distributed, interlinked RDF datasets, treating the web as one queryable database.
Long before “entity SEO” was a marketing buzzword, the original SQUIN was engineering the infrastructure that makes it possible.
The software project has retired. The problem it was solving has not. The intellectual lineage – Linked Data, SPARQL, the Semantic Web stack pioneered by Tim Berners-Lee – is the foundation this publication is built on. The domain wasn’t an accident.
The Entity Grounding Framework
Both consulting offers run on the same methodology: a three-layer model for diagnosing and fixing how AI systems represent a brand.
- Identity layer. Can AI systems identify the brand as a distinct entity, with consistent attributes across the web’s knowledge graph?
- Category layer. Do AI systems place the brand in the correct competitive category, or does positioning collapse into adjacent (lower-tier) categories that quietly route pipeline to competitors?
- Authority layer. Does the brand appear in authoritative sources – Wikidata, G2, Capterra, analyst pages, founder and exec entity records – with enough consensus signal to earn citation?
Every audit answers three diagnostic questions mapped to those layers. Every sprint addresses gaps inside them.
Read the full Entity Grounding Framework →
Our research and technical briefs
Research is published under the Squin Research byline. Edited by Ričerdas Dapkevičius. The publication is organized around three pillars:
1. Semantic SEO
How to optimize for Google’s Knowledge Graph using entities rather than keyword strings. Topical authority, content clustering, entity disambiguation, and the shift from string-matching to meaning-based ranking.
2. Structured Data & Schema Markup
Code-level guides on Schema.org vocabulary, JSON-LD implementation, rich result eligibility, and the structured data types that matter most for visibility in both traditional search and AI-generated answers.
3. Semantic Infrastructure
Analysis of the modern semantic stack – Knowledge Graph APIs, LLM citation trackers, and entity databases. We document what works in enterprise production environments and where the tooling falls short.
Every tutorial is tested against a live implementation before it ships. When Google’s guidelines change, we update the affected pages.
Working with Squin
Two engagements:
- AI Disambiguation Audit. A 25–35 page diagnostic on where your entity signals are breaking and which competitors are getting cited in your place across ChatGPT, Perplexity, Gemini, and AI Overviews. 7–10 business days, plus a one-hour walkthrough with your marketing and exec team.
- LLM Grounding Sprint. A 90-day fixed-scope engagement that fixes the underlying brand signals: schema implementation across templates, Wikidata entity records, knowledge graph signal coordination, and category disambiguation across authoritative sources. Performance benchmarks contracted upfront.
If you’re a B2B SaaS marketing leader and your CEO has asked why competitors are showing up in Perplexity and your brand isn’t, that’s the conversation we have.
Connect
LinkedIn: Ričerdas Dapkevičius
X: @squinorg
Email: [email protected]