Screening Deals on Autopilot with Claude and Realastat

If you underwrite 20 deals to buy one, the bottleneck is not judgment. It is throughput. Most investors run out of time before they run out of deals worth looking at.

The combination of Claude and Realastat changes that. Once your buy box is configured and the MCP connection is active, you can ask Claude to find deals matching your criteria, run full underwriting on each one, and bring back only the ones that pencil. You stay focused on decisions. Claude handles the screening.

Before you start: prerequisites

• Realastat account with at least one analyzed deal in your portfolio
• Claude connected to Realastat via MCP (see our setup guide)
• Buy box configured in Realastat with your target markets, returns, and price range

Step 1: Tell Claude what you are looking for

Open a Claude conversation and describe what you want in plain English. Claude reads your buy box criteria from Realastat automatically, so you do not need to re-specify your targets every time. A simple prompt works well:

Find me 5 rental deals in Indianapolis under $300k that match my buy box.

You can be more specific if you want:

Find me duplexes or triplexes in Columbus OH or Dayton OH under $350k. I want at least 8% cash-on-cash and DSCR above 1.2 at current rates.

Step 2: Claude reads your buy box and searches listings

Claude first calls the Realastat get_buy_box tool to read your saved criteria. It uses those targets to guide its search, so deals that are clearly outside your parameters get filtered before underwriting even runs.

Claude then searches current listings matching your location and price criteria. This step typically takes 30 to 60 seconds depending on how many candidates it is evaluating.

Step 3: Claude runs full underwriting on each candidate

For each property that passes the initial filter, Claude calls the Realastat analyze_property tool. This runs the same AI extraction and deterministic underwriting that you would get by uploading a screenshot manually: cash flow, DSCR, cap rate, cash-on-cash return, break-even rent, and max offer price.

The underwriting uses the regional market assumptions appropriate for each property's location. You get the same calculation quality as a manual analysis, just without the manual steps.

Step 4: Claude surfaces only the deals that hit your targets

After running underwriting on all candidates, Claude filters to the ones that meet your buy box criteria and presents them as a ranked list. Deals that do not hit your minimum cash-on-cash return or DSCR are excluded from the summary.

Claude saves the deals that pencil directly to your Realastat portfolio. You can review them in the dashboard, edit any inputs, and run additional scenarios on the ones you want to explore further.

Step 5: Review in Realastat and go deeper on winners

Open your Realastat portfolio and you will see the deals Claude found, already analyzed and saved. Star the ones worth a second look. From here, you can run scenario analysis, adjust assumptions, calculate your max offer price, and export to Excel for a deeper review.

Other useful prompts to try

• "Compare my top 3 starred deals side by side on DSCR and cash-on-cash return"
• "Run a scenario on the Columbus duplex if I put 25% down instead of 20%"
• "Which of my saved deals has the highest break-even margin?"
• "Screen these 5 addresses and tell me which ones pencil at my buy box targets" (then paste addresses)
• "Update my buy box: I want to increase my minimum cash-on-cash to 9%"

What this does not replace

The autopilot workflow is a first-pass filter, not a substitute for due diligence. Deals Claude surfaces still need property inspections, contractor bids, title searches, and a careful review of the actual rent rolls. The underwriting runs on estimated market data and listing information. Treat the output as a qualified lead list, not a final decision.

What it replaces is the manual step of opening 20 listings, copying numbers into a spreadsheet, and finding out 19 of them do not pencil. Claude does that screening in minutes so you can spend your time on the ones that are actually worth pursuing.