Website AI chat and public-information research
Turn public content
into ready-to-use
AI chat.
<script src=".../widget.js" async></script>USE CASES
FOR LOCAL GOVERNMENT
Import public government and municipal websites to answer questions about services and procedures with links to the source pages.
FOR CONSTRUCTION
Help construction companies and architecture firms research ordinances, guidelines, and filing requirements that vary by municipality, based on project location, intended use, and other details.
This does not automate legal interpretation or decisions about ordinances and regulations. Always check the latest original rules and consult the responsible authority before making a final decision.
PRODUCT PROMISE
Add a script tag to existing HTML to display a self-contained AI chat interface.
Search crawled public pages for information relevant to each question.
Answers and source pages are shown separately, so users can verify the information on the spot.
VISITOR EXPERIENCE
Configure up to six questions per site to help users get started.
Answers appear progressively through SSE, reducing the sense of waiting.
Source-page titles and URLs are displayed separately from the answer.
Ask follow-up questions using the latest 20 history entries for context.
Reset the history to start another question with a fresh search context.
Display Markdown, tables, code, and links, with support for smartphones.
KNOWLEDGE & ANSWER PIPELINE
Set a name, URL, crawl depth, and maximum page count
Respect robots.txt, sitemaps, and Crawl-delay
Retrieve and store titles, body text, and links
Organize the captured information so AI can find it
Save the organized information ready for search
The crawler is designed for large sites, with progress stored in a database-backed frontier so interrupted crawls can resume.
Combine questions in the conversation to form a search query
Search the imported pages for information related to the question
Pass the retrieved context to an LLM to generate a response
Return the SSE answer and source pages separately
If retrieval does not work well, the chat continues using available information or switches to a response that takes safety into account.
ADMINISTRATION & OPERATIONS
Manage display names, URLs, crawl depth, and maximum page counts.
See running, completed, and failed jobs along with error details in one list.
Track titles, URLs, depth, HTTP status, and retrieval timestamps.
Review questions from both the widget and registered users, with details by question, source, user, and date.
Save conversation lists and message details to revisit multi-turn exchanges later.
Add or remove administrators and regular users, change roles, and manage usage status.
FAQ
Explore suitable uses, pre-launch checks, and the decisions that still need human judgment.
It suits websites with public pages such as product information, services, procedures, and FAQs, where users need to move from an answer to its source. It can also support initial research across multiple public websites.
We define the target URLs and crawl scope, then check answers and source pages using expected questions. Before launch, we also verify the widget display, allowed origins, captured pages, and crawl status in the admin console.
Register the target site and crawl scope, import the public content, and add one widget.js script tag to your existing HTML. Check answers, sources, and allowed origins before publishing.
It can find relevant information on registered public pages, such as official municipal websites, and show answers with source pages. Verify applicability and legal interpretation against the latest original text and with the responsible authority.
Use municipal building and development pages relevant to a planned site for initial research into potentially applicable ordinances, guidelines, application instructions, and consultation contacts.
Source-page titles and URLs are shown separately from the answer. Users can open the original source to check its content and update date.
We provide individual guidance after reviewing target sites, page scope, intended use, evaluation questions, and post-launch operational requirements. Include your target URLs and intended use case when contacting us.
IMPLEMENTATION
Set a name, URL, crawl depth, and maximum page count
Retrieve public pages and organize them for AI search
Add the widget.js script tag to existing HTML
Check CORS origins, answers, and sources before launch
CONTACT