# Austin AI Hub > Austin AI Hub is a 501(c)(3) nonprofit in Austin, Texas. Its tagline is "Austin's AI Community for Everyone" and its belief is "AI is being built for people. We help people build it themselves." Everything it offers is free. It exists for the people usually left out of AI: engineers and educators, students and career changers, healthcare workers, small-business owners, parents, and people who have never typed a prompt. ## What Austin AI Hub does - AI Education: free, hands-on AI literacy led by PhD researchers. No coding required. - AI for Good: building public-interest AI with nonprofit partners, including work with the UN on anti-trafficking. - Community: free events and networking for a community of 500+ members, with 600+ attendees across past events. - Mentorship, speaking, and volunteer opportunities for educators and engineers. ## The AI Readiness Check A free tool at https://readiness.austinaihub.org. Given a domain, it fetches that organization's home page once, reads its robots.txt and llms.txt if they exist, and looks up five DNS records. From those eight public requests it infers what kind of organization it is (nonprofit, school, clinic, government body, retailer, professional firm, software company, or local business), names the systems visible from outside, reports which AI crawlers the robots.txt allows, scores AI readiness out of 100, and recommends free Austin AI Hub programs that fit. - It does not crawl. It fetches exactly one page and never follows a link, so a system that appears nowhere on the home page appears nowhere in the results. The findings are a floor, not an inventory. - It never sees anything behind a login, and it runs no port scan or intrusive probe. - Nothing is stored. There is no database, no account, and no follow-up. - It works for any organization, not only one industry. Guidance, privacy obligations, and example workflows change to match the organization type. - When a DNS lookup or a file request fails, the check marks that item unread rather than scoring it against the organization. An absent SPF record and a timed-out lookup are different claims. - Scores and hour estimates are illustrative, not a promise or a quote. ## What is AI readiness? AI readiness measures whether an organization can put an AI tool to work on its real information. It covers where records live, who may read them, whether anyone counts results, and whether machines can parse the organization's public pages. The score predicts how much groundwork a first AI project needs. It predicts nothing about revenue. ## How the score is calculated Seven weighted components, published rather than hidden: | Component | Points | What it measures | | --- | --- | --- | | Machine legibility | 18 | Crawler access, structured data, llms.txt, page structure | | Foundations | 18 | A shared workspace and modern hosting | | System of record | 15 | One dependable place contacts and history live | | Security and privacy | 15 | Web headers and email authentication, graded from public sources | | Measurement | 12 | Whether the organization already counts what happens | | AI experience | 12 | Evidence that someone has shipped AI to real people | | Connectedness | 10 | How many separate systems hold a piece of the picture | Connectedness uses diminishing returns: an organization running fifteen tools is not three times readier than one running five. Scores clamp between 12 and 96, because nothing visible from the outside justifies either extreme. ## How the detection works - 586 vendor fingerprints across 37 categories. - Fingerprints match four evidence surfaces: the URLs a page loads, HTTP response headers, cookies, and the host allow-list in the Content-Security-Policy header. - Fingerprints never match visible prose. A page that names a vendor as a customer does not run that vendor. - DNS-over-HTTPS resolves A, NS, MX, TXT and DMARC records to identify the DNS host, the mail provider, SPF senders, and SaaS tools proven by domain-verification tokens. ## AI crawler policy, and why it matters The check reads a site's robots.txt and reports which of 17 named AI crawlers are allowed, split by what each one is for: - Search crawlers (10): OAI-SearchBot, ChatGPT-User, Claude-SearchBot, Claude-User, PerplexityBot, Perplexity-User, Bingbot, Amazonbot, DuckAssistBot, MistralAI-User. Blocking one of these means that assistant cannot cite the organization, even when it is the best answer. - Training crawlers (7): GPTBot, ClaudeBot, Google-Extended, Applebot-Extended, Meta-ExternalAgent, CCBot, Bytespider. Blocking one of these keeps the organization's text out of the next model and costs it no citations. Most robots.txt files that block AI do not make this distinction, and the organizations that wrote them are invisible to answer engines by accident. ## How the Hub frames the journey - Learn (start this month): build shared understanding before anyone touches a tool. Free workshops and community events. - Build (next one to three months): connect the systems you already run, then automate one real workflow. - Lead (when you are ready): partner on AI for Good projects, volunteer, and teach what you learned. ## Key pages - Main site: https://austinaihub.org - AI Readiness Check: https://readiness.austinaihub.org/ - Programs detail: https://readiness.austinaihub.org/programs.md - Pricing: https://readiness.austinaihub.org/pricing.md ## Contact - Organization: Austin AI Hub, a 501(c)(3) nonprofit - Address: 5900 Balcones Dr STE 100, Austin, TX 78731 - Email: team@austinaihub.org - Cost: everything is free. The Hub sells nothing.