skan.ai — AI Site Grade

Analysis

The Cold-Knowledge Gap Is the Story

Cold LLM knowledge describes Skan as a "process discovery" platform founded in 2019 by Avinash Misra and Sanjay Bhasin, with $40M+ funding led by Sequoia Capital India. The actual site contradicts every one of those details — the founding date, the co-founder names, the investors, and the product positioning have all shifted significantly.

Schema Posture

JSON-LD is present on every page examined, using Organization, Corporation, WebSite, WebPage, Product, SoftwareApplication, BlogPosting, TechArticle, and NewsArticle types. The Organization schema declares a foundingDate of 2016-01-01, but the About page timeline says "Skan AI Founded September 2018." The cold LLM says 2019. Three different founding dates across schema, site copy, and external knowledge — a fragmentation that erodes trust signals for any AI engine doing entity resolution. The Product schema on /platform uses SoftwareApplication with applicationCategory: "BusinessApplication" and lists price as "0" with a note "Contact for pricing," which is technically valid but semantically thin. No FAQPage or HowTo schema exists anywhere, despite the site using FAQ-like question headings in blog posts.

Crawler Access

The robots.txt is unusually sophisticated: 24 user-agent rules explicitly allow every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Applebot-Extended, anthropic-ai, cohere-ai, YouBot, PhindBot) while blocking CCBot, omgili, and webzio-extended. compare_bot_access confirms every AI bot tested returns 200 with identical 265KB payload — no UA-based blocking, no Cloudflare challenge, no JS shell. This is best-in-class AI crawler access posture. However, /llms.txt returns a 404 (HubSpot error page), meaning there is no structured AI-friendly content map. The sitemap.xml is present with 418 URLs and is well-organized.

Cold-Knowledge Gap

The cold LLM believes Skan was founded in 2019 by Avinash Misra and Sanjay Bhasin, with a $30M Series A from Sequoia Capital India in 2022. The site says founded September 2018 by Avinash Misra and Manish Garg (not Sanjay Bhasin), with a $40M Series B in March 2022 from Cathay Innovation, Citi Ventures, Dell Technologies Capital, Firebolt Ventures, and Zetta Venture Partners — no Sequoia. The cold model also describes the product as "Process Intelligence Suite" focused on task mining integrated with RPA tools like UiPath. The site now positions itself as the "Context Graph of Work" and "Agentic AI Platform" — a significant repositioning toward agentic AI that the cold knowledge has not absorbed. The cold model also references "over $40 million in funding" which roughly matches, but the investor mix is completely different.

External Signals

The site references a Gartner Magic Quadrant for Process Intelligence Platforms 2026 honorable mention, a World Economic Forum New Champions membership, and an acquisition of Metaculars Inc. (October 2025). DNS TXT records show verification tokens for anthropic, apple, cursor, notion, and parallels — indicating active AI-toolchain integrations. However, web searches for external reviews, Reddit discussions, and third-party analyst coverage return near-zero results, suggesting limited off-domain footprint. The case studies are all anonymized ("F500 Financial Services," "F300 Benefits Provider") with no named logos, which reduces citation value for AI engines seeking verifiable claims.

Findings

1. Founding date fragmented across schema, site, and external knowledge High

   The Organization schema declares a foundingDate of 2016-01-01, the About page timeline says September 2018, and cold LLM knowledge says 2019. This inconsistency erodes trust signals for AI entity resolution.

   What to change: Align the foundingDate in Organization schema with the site's stated founding date (September 2018) and ensure consistency across all pages.

2. Missing /llms.txt file High

   The /llms.txt endpoint returns a 404 (HubSpot error page), meaning there is no structured AI-friendly content map for LLMs to discover key pages and context.

   What to change: Create an /llms.txt file that lists the most important pages (e.g., /platform, /about-us, /pricing, /knowledge-hub) with brief descriptions.

3. Cold LLM knowledge outdated on product positioning High

   Cold LLM knowledge describes Skan as a 'Process Intelligence Suite' focused on task mining with RPA integration, but the site now positions itself as the 'Context Graph of Work' and 'Agentic AI Platform'. This gap means AI engines may misrepresent the product.

   What to change: Publish a press release or update the company's Crunchbase, Wikipedia, and other authoritative sources to reflect the new positioning and funding details.

4. Cold LLM knowledge lists wrong investors High

   Cold LLM knowledge claims Sequoia Capital India led a $30M Series A, but the site states a $40M Series B led by Cathay Innovation, Citi Ventures, Dell Technologies Capital, Firebolt Ventures, and Zetta Venture Partners. No Sequoia involvement is mentioned.

   What to change: Update external profiles (Crunchbase, PitchBook) with accurate investor information and consider publishing a funding announcement on the site.

5. Near-zero external signals from reviews, forums, and analyst coverage Medium

   Web searches for reviews, Reddit discussions, and third-party analyst coverage return zero results. This limits off-domain footprint and reduces the ability for AI engines to find corroborating citations.

   What to change: Encourage customer reviews on G2, Capterra, and TrustRadius; engage in relevant Reddit and Quora discussions; and pitch to industry analysts for coverage.

6. Case studies anonymized with no named logos Medium

   All case studies use generic descriptions like 'F500 Financial Services' without naming the customer. This reduces citation value for AI engines seeking verifiable claims.

   What to change: Seek permission to use customer logos and names in case studies, or add testimonials with named individuals.

7. Product schema lists price as '0' with no meaningful pricing info Low

   The SoftwareApplication schema on /platform has price '0' and a note 'Contact for pricing'. While technically valid, this is semantically thin and may confuse AI engines expecting a real price or 'Contact for pricing' as a string.

   What to change: Use a priceSpecification with price '0' and a description 'Contact for pricing', or omit the price field if not applicable.

8. No FAQPage schema despite FAQ-like content Low

   Blog posts use FAQ-like question headings but lack FAQPage schema. This misses an opportunity to appear in AI-generated FAQ snippets.

   What to change: Add FAQPage schema to blog posts that contain question-answer sections.

What's working

• Best-in-class AI crawler access with explicit allow rules — Robots.txt explicitly allows all major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, etc.) and compare_bot_access confirms 200 status with identical payload for all tested bots. No UA-based blocking or JS shell.
• Comprehensive JSON-LD schema across all pages — Every examined page includes JSON-LD with types Organization, Corporation, WebSite, WebPage, Product, SoftwareApplication, BlogPosting, TechArticle, and NewsArticle. This provides rich structured data for AI engines.
• Active AI-toolchain integrations via DNS verification tokens — DNS TXT records show verification tokens for anthropic, apple, cursor, notion, and parallels, indicating active integrations with AI tools and platforms.
• Well-organized sitemap with 418 URLs — Sitemap.xml is present and contains 418 URLs, providing a clear content map for crawlers.
• Knowledge hub with extensive content (12,775 words) — The /knowledge-hub page contains 12,775 words of AI knowledge management and enterprise automation resources, providing substantial material for AI engines to index.
• Acquisition of Metaculars Inc. published on site — The site has a news article about acquiring Metaculars Inc. in October 2025, providing a verifiable external signal of company growth.