Report chat is an analyst you can talk to. Ask a question about the office market in everyday language and it finds the numbers, shows them as a chart or a table, and turns the conversation into a report you can download.
Step by step
Something like “How has vacancy moved since 2021?” There is no special wording to learn and no menu to find.
It works out where the answer lives in the market database and looks it up. If a lookup goes wrong, it reads what went wrong and tries again.
Charts and tables appear in the conversation as it goes. You can open any step to see exactly what it looked at.
Before headline figures go into a report, a separate checker works them out again on its own. If the two disagree, you are told.
Ask for a report and download it as a PDF or a PowerPoint deck, built from the charts and findings in the chat.
Judgment
These are the moments a simple search tool gets wrong. The demo data is built to include them.
Net rent or gross, a quarter or a year, one submarket or the whole metro. It picks the one that fits, and says which it used.
A trend becomes a line chart, a ranking becomes a table. It chooses the view that makes the point clearest.
If the numbers do not support a conclusion, it says so rather than filling the gap with a guess.
Who does the work
A large, capable model that plans the work, looks up the numbers and writes the answer.
A smaller, faster model that recalculates the key numbers from scratch before they reach a report.
Every agent here is built the same way: a leading AI model that plans the work and a smaller one that checks it, a set of tools each is allowed to use, and rules on what it must prove before it answers. They run on leading models from Anthropic or OpenAI, inside this website.
Safeguards
Three invented cities (Harbor City, Pinegate and Riverton), twelve submarkets and five years of quarterly figures. The numbers are built to agree with each other, so a chart never contradicts the table beside it.