CRM and customer data 3 min

A CRM that fills itself as you talk: the Palmora Property case

An AI agent connected to the CRM fills in records instead of your teams: they talk or write to a bot, which creates or updates contacts, companies, deals and listings. At Palmora Property, agents save at least 60% of their writing time, the CRM is filled because nobody has to open it, and proofreading stays human.

Contents6
  1. 01The real problem with a CRM is data entry
  2. 02At Palmora Property: a Telegram bot connected to the CRM
  3. 03What it changes
  4. 04Why Telegram rather than WhatsApp
  5. 05What stays human
  6. 06Our view

“Talking to a machine will become the norm.” The line is everywhere now that synthetic voices have become stunning. It is right, but it misses the point. Voice is the shop window. What matters is what the machine does with what it hears, and above all where it files it.

The real problem with a CRM is data entry

A CRM is only worth the freshness of its data. And everyone knows it: sales reps and agents do not fill it in, or late, or halfway. Not out of bad will, but because opening the tool, finding the record and typing a report after each visit takes time they would rather spend with clients. We cover this in why sales reps do not fill in the CRM.

The fix is not to chase them. It is to remove the data entry.

At Palmora Property: a Telegram bot connected to the CRM

Palmora Property is a real estate matchmaking platform in Phuket. We set up a Telegram bot for its agents, connected directly to the CRM and the website. In practice, an agent:

  • creates a listing by sending its details and photos to the bot, which writes the page and prepares it for the site;
  • writes and publishes property and client records more easily, without copy-pasting between ChatGPT and the back office;
  • updates a deal, a contact or a company by talking: they leave a voice note after a visit, the bot transcribes it and files each piece of information in the right record.

From a voice note to an up-to-date CRM

  1. The agent talks

    A Telegram voice note, right after the visit.

  2. The bot transcribes

    It identifies the contact, listing and deal involved.

  3. The CRM updates

    Each piece of information goes in the right field, overwriting nothing.

  4. The agent reviews

    They approve the record before anything is published.

They open nothing. They talk, between two meetings, and the CRM is up to date.

What it changes

The first gain is time: agents save at least 60% of the time they spent writing. And they already used ChatGPT before, with copy-paste: the gain comes from removing the back-and-forth, not just from AI.

The second gain is less visible and more valuable: the CRM is finally filled in. Since updating it no longer takes effort, agents do it, right away. The data is more complete and fresher, and the tool is truly adopted. They focus on what matters: the client, the visit, the negotiation.

Why Telegram rather than WhatsApp

We are often asked, since everyone has WhatsApp. For an internal team automating its work, Telegram is far better suited:

  • a bot takes minutes to create, for free, with no business verification or professional account to get approved;
  • no per-conversation billing and no message templates to get approved, unlike the WhatsApp Business API;
  • useful features out of the box: buttons, menus, large file sending, groups and topics, voice notes a bot can easily work with;
  • an open, stable API, designed for automation.

WhatsApp remains the right choice for talking to your customers, who already have it: that is the subject of our article on the WhatsApp Business chatbot. But as your teams’ work tool, Telegram is faster and cheaper.

What stays human

The bot has not replaced the agents’ work, it has removed its most tedious part. Three things stay in their hands:

  • proofreading each record before publication;
  • a test after publication, to check the listing displays as intended;
  • feedback on errors: at first, the bot sometimes got things wrong. The agents reported them to us, and each report was used to fix the automation. That is what made it reliable.

One simple rule prevents damage: the bot only fills a field if the conversation gave it the information. It completes the record, it never overwrites what is there.

Our view

Voice is solved, any tool can transcribe and speak. The real question for a business is what comes next: does your CRM know what to do with what the machine hears? If not, the voice agent stays a gadget. If so, data entry disappears, and with it the main reason CRMs are poorly kept.

Jim · founder of Brain Plus AI

Also in crm and customer data

  • Understand6 min

    CRM adoption: why salespeople don't fill it in

    When a CRM stays empty, it is almost never ill will: the tool asks salespeople for more than it gives back. The reasons that keep coming up, and what makes a CRM the tool the team opens of its own accord.

  • Practical guide8 min

    Managing listings and mandates in a CRM: the data model

    A useful real estate CRM keeps properties, mandates, owners, buyers and their searches in one tool, connected to the listings website. What it should contain, the French legal obligations it helps you meet, the example of a portal we run on Twenty, and its limits.

  • Understand7 min

    AI voice agents on the phone: what they can do in 2026

    A voice agent picks up, understands a spoken request and answers with a synthetic voice, connected to your phone system. What it does well in 2026, what it does badly, the rules to follow, and the place of the people who answer the phone.

WhatsApp