SuiteCRE
Start free trial
Menu

AI & agents

AI that drafts the work. You decide what goes out.

Every AI feature in SuiteCRE reads real data from your account and hands you work you can check. Nothing reaches a client until you approve it, which is deliberate.

Start your free trialHow it works

The rule everything follows

Drafts and proposals, not surprises

The fastest way to lose a client is software that emails them on your behalf. SuiteCRE is built so anything with an outside consequence lands in your queue first.

Approval required

Emails, messages, and calendar invites to other people are produced as drafts. You review, edit, and send. Drafts hand off to your own mail client, which is why SuiteCRE never asks for permission to send as you.

Automatic, but internal

Creating a task or logging an interaction can run on its own, because nothing leaves the account. Every one is written to an audit trail you can review.

The one exception

Buyer chat speaks to prospects directly. It works under tight limits: it answers from your documents, refuses to negotiate, and escalates the categories you flag.

What it does

Five surfaces, one assistant

They read the same data, follow the same rules, and hand work between each other. Core CRM includes a limited assistant and buyer chat; the full assistant, the agents, and the automations are on Broker AI.

01

The assistant

It has already read your CRM

The assistant answers against your contacts, deals, listings, documents, calendar, and commissions. Tag a record with @ and the answer is about that record, not a general one.

  • Ask about pipeline state, stale relationships, or what a lease actually says
  • Search your documents and get the passage back with its source cited
  • Draft emails, follow-ups, listing copy, and note summaries
  • Paste in a screenshot, flyer, or photo of a page and ask about it
  • Create tasks, notes, logged interactions, and calendar events
  • Live voice for when you are driving between showings
Read the documentation
02

Buyer chat

Your listings answer questions at 11pm

A public chat page attached to a listing. It answers from the offering memo and supporting documents you uploaded, and hands anything sensitive back to you with the prospect's details captured.

  • Answers from the listing record and its documents, without speculating or inventing numbers
  • Ten escalation categories on by default, each one you can turn off
  • Your own keywords on top, plus control over what happens to unanswerable questions
  • Stays off negotiation, valuation opinions, and seller motivation
  • Cannot reach your document library, your other listings, your contacts, or your deals
  • Every conversation stored, with filters for what it handled and what escalated
Read the documentation
03

Revenue agents

Five specialists working your pipeline

The assistant waits to be asked. Agents run on their own schedule. Acquisition, Deals, Due Diligence, Financing, and Lease Admin each watch a different part of the business and bring you work.

  • Acquisition runs your saved searches and queues buyer-matched opportunities
  • Deals drafts follow-ups and suggests structures grounded in that deal's own numbers
  • Due Diligence tracks deals toward closing and flags gaps and stalls
  • Lease Admin watches dates, options, and reconciliation obligations after the close
  • Work arrives as action cards with the reasoning and the underlying data attached
  • Run any agent on demand, or turn one off, in which case it stops instead of running quietly
Read the documentation
04

Your own automations

Describe it in plain English, review it, turn it on

Agents are ours; automations are yours. Describe a recurring instruction in the workspace builder and get back a recipe with a trigger, a condition, and a set of actions.

  • Recipes arrive paused, so you read what one will do before it does anything
  • Triggers for new contacts, new call notes, deals gone quiet, and daily checks
  • Task creation and interaction logging can run automatically
  • Email drafts and calendar actions stay approval-bound, always
  • Every action validated against the approved action registry before it executes
  • A full audit trail of what ran, when, and why
Read the documentation
05

Anywhere you are

WhatsApp, SMS, and Claude

The same assistant and the same data, reachable from wherever you happen to be working. A note added from a parking lot is the same note on the deal in the web app.

  • Pair your phone number and text the assistant over WhatsApp or SMS
  • Ask for status, make changes, dictate what happened on a call
  • Long requests run in the background instead of timing out
  • A daily brief at a time you choose, which is the part most brokers use every day
  • Connect your account to Claude and work against your CRM in any conversation
  • Identical approval rules on every channel: it drafts, you send
Read the documentation

Bring your own Claude

Connect SuiteCRE to Claude

SuiteCRE publishes a custom connector for Claude. Connect it once and you can open Claude on the web, desktop, or your phone and work against your own CRM, reading your pipeline and saving drafts back into SuiteCRE.

For example

“Who are my top ten leads this month? For each, draft a personalized second-touch email referencing our last interaction, and save the drafts.”
“What does the lease for 123 Main St say about the renewal option? Add a note to that deal summarizing it.”

A connection only ever reaches your own records, and it is read-only unless you grant write access. Write access saves email as paused drafts, so Claude cannot send mail on your behalf either.

How to connect it

Straight talk

Where you still have to do the work

We would rather tell you this now than have you find it out in front of a client.

Verify anything client-facing

Lease abstracts, extracted figures, and market data are a reading aid, not a substitute for review. Every extracted field points back at its source so checking it takes seconds. Do the checking.

Data quality varies by market

CRE public records are uneven. Recorded prices can be nominal, ownership sits behind entities, and rent data is thin in secondary markets. The platform flags what looks unreliable rather than smoothing it over.

It is only as good as your uploads

The most common cause of a wrong answer is a document that was never uploaded or a note nobody wrote. The assistant reads what exists and knows nothing about what you kept in your head.

See what it drafts for your pipeline

Import your contacts, add one live deal, and ask the assistant something you actually want to know. That is a fair test in about ten minutes.

Start your free trialRead the docs

The documentation spells out exactly what each AI feature can reach and what it cannot.