A small data footprint.
MapLedger is operated by Dark Vector Cognition. It requires no account. Saved businesses and your comparison stay in this browser’s local storage; they are not synced to a server. Exports are generated on request and downloaded to your device.
Search terms are sent to our hosting service to retrieve matching public records. Hosting infrastructure may record request metadata and IP addresses in operational logs. We do not add advertising trackers or sell visitor profiles.
Paid territory packages
Payment is handled by our payment provider; we never receive your card details. We retain the confirmed scope, package, payment status and receipt email for up to 400 days after payment. An unpaid package expires after 30 days to allow delayed payment confirmation. Your download page has a private, unguessable link. Keep it safe: anyone with the link can access the files. Private order links are excluded from page-view analytics.
When you arrive through our territory campaign, a cookie namedml_campaign remembers that fixed campaign label for 30 days. It contains no person or order identifier. We attach the label to your order to count checkouts, confirmed payments and download responses. A download response does not prove that you saved or used the file. Global Privacy Control or Do Not Track disables this measurement and clears the campaign cookie. Campaign totals expire after 45 days. Test transactions are separate from customer sales; duplicate-event flags expire after 400 days.
What we measure
We use Vercel Web Analytics, which sets no cookies and does not follow you across sites, to count page views and a few product actions: a submitted search and whether it found anything, adding a business to your shortlist or comparison, and starting an export. These events carry only fixed choices such as the sector filter, the export format and a rough result count. Your search words, the businesses you pick and anything you type are never sent as analytics.
Our own server also counts each page view, with no third party in the path, and sets one cookie, dvc_v: a random id, kept for a year, that tells a returning browser from a new one and carries nothing about you. For each day we keep totals by page, by the site you came from, by campaign tag, by country and by device type, with known crawlers counted apart. Unique visitors are estimated from a hash of your address and browser that is salted per day, so it cannot be joined from one day to the next; we keep the hash, not the address. We also keep a short label for where a new browser's first visit came from, such as "search" or "reddit", so we can count where visitors and Lot File orders come from. That label, and the day of that first visit, are also written to the payment session when you buy a Lot File. If your browser sends Global Privacy Control or Do Not Track, this records nothing and sets no cookie. Everything here expires after 400 days.
On the Nashville land and zoning map we count which ready-made view is chosen, whether an address lookup found a match, and whether a printable brief was opened, with fixed choices only. The address you type is looked up inside your browser against Metro Nashville's street list and is never sent to us or to a geocoding service. The spot you pick is sent to our server, which asks Metro Nashville's public map services for that lot's land use, building age, street frontage, service district and rental permit counts, so the rental answer is the lot's own. We do not store the spot, and we never read owner names. A printable brief carries the place in its web address; before a page view is counted, every part of the address except the chosen map view and campaign tags is removed.
On the Austin land and zoning map we count the same kinds of events. There, the address you type and the spot you pick are sent to our server, which asks the City of Austin's public map services for the matching address point, the lot's size, the zoning ordinances and the land use the City records at that spot. The typed address travels in the body of the request, not its web address, and we do not store either. The printable brief is redacted the same way as Nashville's before a page view is counted.
Buying a Lot File. Stripe takes the payment; we never see your card. To prepare and deliver the file we keep the order for about a year: the lot's point, the address you typed, the email Stripe gives us (to send you the link), whether and how much you paid, what the research cost us, and the file itself. Our research agent reads Metro Nashville's public records about the lot; the plan documents are read by Anthropic's model, which receives only those public documents and the plan's identifiers, never your email, your name or the address you typed. The order page is reachable only through its unguessable link. On the map we count that the offer was shown and that the buy button was pressed, with fixed choices only.
If your browser sends Global Privacy Control or Do Not Track, the analytics script is not loaded and no product events are sent. The tool works the same either way.
Map tiles load from OpenStreetMap, and fonts load from Google Fonts. Those providers receive the network information necessary to serve their resources. External website and social links open the destination you choose.
Nashville zoning, permit, tax and flood information comes from Metro Nashville's public records and Metro Council's ordinance text, retrieved on the date shown on each page. Austin zoning, ordinance, flood and boundary information comes from the City of Austin's open data, and lot sizes from the Travis and Williamson appraisal districts' parcel maps as the City publishes them. Business details come from Overture Maps and its named source datasets. The catalogue contains public business contact information, not verified owner identities. Use source links to confirm or correct information at its origin.
Remove a saved business with its star button, or clear this site’s browser storage to remove the entire local list.
MapLedger in Claude, ChatGPT and other assistants
The MCP server at mapledger.app/mcp needs no account and answers from public records. Your assistant sends it only what a request needs: a city, an address or a point, a search, and optional tax inputs. It answers from the same data as the maps and asks the City of Austin's and Metro Nashville's public map services about the lot at that spot. We do not store what was asked.
To count use, each call records which tool was called, the city (Austin, Nashville or another), whether it answered, how long it took in broad bands and a recognized assistant app name (otherwise unknown), never the address, point, search or the other city's name. That record sits in our host's (Vercel) runtime logs for no more than 30 days and is counted in Vercel Web Analytics without forwarding your IP address or cookie. We retain aggregate daily reports to understand use over time. Marked internal checks are reported separately; unmarked calls are not proof of a person. If your assistant sends Global Privacy Control or Do Not Track, no usage event is recorded.
We never receive your conversation, your name or your account. Links in its answers carry a tag that says they came from an assistant. To stop, remove the connector from your assistant. Your assistant's provider handles your conversation under its own privacy terms.
Help and corrections
Questions, data corrections and removal requests go to hello@darkvectorcognition.ai. A business listing can also be corrected at its source through Overture Maps; the next MapLedger refresh picks up source changes.