Talqui Helpdesk: one source of knowledge for your customers, your team and your AI

Most companies treat their help center as a website. A place where customers go to read FAQs, usually hosted on a separate tool, written by someone who isn't in the support queue, and updated whenever someone remembers.

We think that's a waste of the most valuable thing a support operation produces: its answers.

Today we're launching Talqui Helpdesk. Yes, it's a help center your customers will love. But we built it to be something bigger: a single source of knowledge that your customers, your agents, your AI agents and your new hires all draw from, and that you can write and maintain with the AI tools you already use.

1. More than a help center

Let's start with what everyone expects, because it matters.

With Talqui Helpdesk, you write articles in a visual editor inside Talqui, organize them in categories, and publish them as a public help center with your brand: your colors, your logo, your favicon, and your own domain (something like help.yourcompany.com). Customers get a clean site with search, articles that look good on any device, and pages built to be found on Google, with structured data, a sitemap and a social image generated for each article.

That alone replaces a third-party tool many companies pay for separately, and brings the content back to where the support actually happens.

But if we stopped there, Helpdesk would be one more help center. What changes the picture is who else reads it.

2. One source of knowledge, four audiences

A good article answers a question once. The problem is that, in most companies, only customers ever see it. Talqui Helpdesk puts that same article in front of everyone who needs it.

Your customers

Anytime, on any device, without opening a ticket. The best support conversation is the one that didn't need to happen because the answer was already a search away.

Your agents, during the conversation

Inside the Talqui conversation box, there's an Articles tab listing every published article, grouped by category. The agent finds the right one and inserts it into the reply in one click. No switching tabs, no hunting for links in a shared doc, no retyping the same explanation for the tenth time today. The customer gets a complete, well-written answer, and keeps a link they can come back to.

Your AI agents

This is where the knowledge starts working on its own. Talqui's AI agents read your help center to answer customers directly, using the details in your articles instead of generic responses. Every article you publish makes your AI more accurate, and every accurate answer is one less conversation that needs a human. It's the same content, working twice.

Your team

New agents need to learn how your product, your policies and your processes work. Experienced ones need to check a detail quickly. Instead of a scattered wiki, they have the same source your customers and your AI use, which also means everyone is working from the same version of the truth.

3. Written with AI, through MCP

Keeping a help center up to date has always been the hard part. Writing takes time, screenshots get outdated, and the people who know the answers are busy answering customers.

That's why Talqui Helpdesk speaks MCP (Model Context Protocol), the open standard that lets AI assistants use external tools. Connect your help center to Claude, ChatGPT or any assistant that supports MCP, and you can simply ask it to:

• write a new article about a feature, a policy or a process;
• update a specific section of an existing article, without touching the rest;
• create and organize categories;
• look up what your help center already says about a topic.

Everything happens in plain markdown, so the assistant writes the way it naturally does, and Talqui turns it into a published article.

How we use it at Talqui

We've been running our own help center this way, and it changed how we handle support.

Whenever a customer asks something and we notice our help center doesn't have a complete answer yet, especially when it's a question we've heard before, we don't just answer and move on. We create the article right then. An AI assistant connected through MCP writes it, and a capture tool we built produces what the article needs to be truly useful: cropped screenshots, GIFs and even a narrated video showing exactly how the feature works. Then we send the link to the customer.

The customer gets a better answer than a typed reply could give. And the next person with the same question, whether they search on their own, talk to an agent or talk to our AI, finds it already answered. Every new question reduces the chance of having to answer it again.

You can see the result at ajuda.talqui.chat, our own help center (in Portuguese, for our customers in Brazil), built and maintained with the help of this MCP.

4. The layer between your customers, your team and your AI

This is the last launch of our Launch Week, and it closes a story.

On Monday, WhatsApp to Web Chat gave companies a way to keep WhatsApp as the entry point without paying for every message after it. On Tuesday, Workflows let companies make Talqui follow their own process instead of adapting to ours. Today, Helpdesk turns what your team knows into something your whole operation can use.

Put together, they say what Talqui is becoming. Not a chatbot measured by its resolution rate, and not only a customer service platform. We're building the communication backbone of your company: the layer that sits between your customers, your operations team and your AI agents, where messages flow across every channel and knowledge is created, shared and reused. A hub where you plug in your channels, your systems, your internal MCPs and your AI, and run conversations at scale.

Talqui Helpdesk is available now. If you want to set up your help center, move your existing articles, or see how it can power your AI agents, talk to us.

Thanks for following Launch Week #01. This was just the first one.