Use Ploti from the assistant you already use

The AI assistant you already use can work with Ploti. That includes a chat assistant such as Claude, ChatGPT or Gemini, a coding agent such as Claude Code, Codex or Cursor, and an always-on assistant such as Muse. Once connected, it answers real estate questions from official county and city records, such as parcels, zoning, permits and ordinances, with the source attached. It can also ask Ploti to collect data that’s missing. Setup takes a few minutes, and the steps depend on the kind of assistant.

​

The data is in your coverage areas

Ploti collects official public real estate records from US county and city offices and keeps them current. When you sign up, you pick the counties you work in, called your coverage areas. Ploti is priced per county, and your first county is free. Your assistant can read everything Ploti holds for those counties, and Ploti keeps adding data there on its own. Asking Ploti from your assistant uses no credits. Coverage areas and pricing has the prices. When a question needs data Ploti doesn’t have yet, your assistant can file a research request. Ploti collects the missing records from the official sources and emails you when they’re in. Research requests explains how they work.

​

Why connect Ploti

An assistant on its own searches the web. It stalls on county permit portals and scanned PDFs. It has no county records in its training, so it searches again for every question. With Ploti connected, it reads records Ploti has already collected and keeps current, and it names the official source in every answer.

​

What assistants do with Ploti

Two examples of the work an assistant does, beyond answering a question in chat:

• A coding agent doing site selection. “Find parcels in Orange County, FL over five acres, zoned industrial, outside flood zones, held by the same owner for ten years or more.” The agent searches Ploti’s datasets, reads the matching rows, checks what the zoning code allows, and saves the parcels on a map the analyst opens. No copy of the data is needed.
• Showing a person the answer. An assistant puts the qualifying parcels on a map beside the zoning layer and shares a link, so the client sees the answer instead of a spreadsheet.

Show results on a map explains how an assistant saves and shares a map. Example prompts has more.

​

Three kinds of assistant

A connector is an add-on that lets a chat assistant look things up in an outside service. Ploti’s connector uses MCP (Model Context Protocol), the standard that Claude, ChatGPT and Gemini use for connectors.

​

Chat assistants

You chat; Ploti does the work. Your assistant finds the records, Ploti runs any analysis on its own servers, and the result comes back as an answer with sources, a table, or a Ploti map: a link that opens the results on a map in Ploti. How the connector works: Your assistant’s own plan may need to be a paid one, or have a developer setting turned on, to add a custom connector. Each assistant’s page says what it needs.

​

Coding agents

A coding agent runs on your computer and writes and runs code for you. With Ploti’s command-line tool or Python library, it downloads the records it needs and works on them there. It might build a screening script or a data pipeline. Add the connector as well when you want it to save results on a Ploti map you can open.

​

Always-on assistants

These assistants keep running when you are not chatting. They live in your messaging apps, run tasks on a schedule, and message you when something changes. Muse, Grok Bot, Instinct and dots are examples, and an agent you build and run yourself belongs here too. Most are set up by sending one chat message with a link to Ploti’s setup instructions. They then use Ploti’s command-line tool or its web API. Always-on assistants lists each one.

​

Next steps

Building software on Ploti’s data rather than setting up an assistant? See Developers.