Partner Story ยท in conversation with Teydra Lab. This is a written interview submitted through Submit Your Story. The claims below are the company's own, presented as they were given to us. No payment was involved.
There is a quiet arithmetic problem sitting under most website chatbots, and it has nothing to do with the model.
You buy one. It costs a monthly fee. The fee runs whether the bot answered four hundred questions in November or none in February. Traffic grows, so the tier goes up. You want a different tone of voice, a new answer about your return policy, a colour that matches your brand, and now you are asking someone else to change it, waiting for them to have time, and often paying again for the privilege.
Teydra Lab, a platform operating out of Salzburg, Austria, has taken a swing at that arrangement from an unusual angle: charge once, then get out of the way.
What the product is
Teydra Lab is a configurable AI chatbot for websites. The pitch is not that its answers are smarter than anyone else's. It is that the person who owns the website also owns the controls: what the bot knows, how it speaks, what it looks like, which languages it handles, and which model sits underneath it.
The onboarding is built to demonstrate that before any money changes hands. You give the platform a website address, answer some questions, and look at a draft bot built from your own content. No account, no demo call, no sales conversation standing between a visitor and the thing they came to evaluate. An account is only required when you decide to keep it.
Configuration and internal preview stay free. What the payment unlocks is public deployment: the widget on your live site, the public direct link, the QR code.
The money, which is the interesting part
The unlock is a single payment of 237 euros, presented as an introductory price reduced from 949 euros. There is no monthly fee after it. What continues is usage: AI costs are billed through a token balance the owner loads, with consumption tiers carrying fixed multipliers depending on how heavy the model and the task are, and automatic top-ups switched off unless the owner explicitly turns them on.
Read that structure carefully, because it inverts the usual incentives. In a subscription model, a chatbot nobody uses still bills you, and a chatbot that becomes popular bills you more through tier upgrades that have little to do with your actual compute. Here, a quiet month costs close to nothing beyond the initial unlock, and a busy month costs whatever the conversations genuinely consumed.
For a small firm, that difference is not philosophical. A dental practice, a regional installer, a small hotel: these are businesses where a chatbot answers the same fourteen questions endlessly, and where a recurring line item gets cancelled the first time somebody reviews the software budget. Removing the recurring line removes the moment where it gets cancelled.
The trade is honest and worth stating plainly. One-time pricing means the vendor's revenue from a customer largely arrives up front, which is excellent for the customer and demanding for the vendor. It works only if the product genuinely does not need constant hand-holding. That is exactly what the company is betting on, and it explains why so much of the product is aimed at letting owners change things themselves rather than filing a request.
Compliance built in, not bolted on
The part that caught our attention as a publication is smaller than the pricing and, in our view, more significant.
Consent handling, the operator's own privacy texts and EU AI labelling are described as built into the widget rather than sold as an upgrade. The platform also publishes a data processing agreement alongside its terms and privacy policy, which is the document any European company needs on file before it can legally let a third party process customer conversations on its behalf.
Since 2 August 2026, Article 50 of the EU AI Act has required that people be told when they are interacting with an AI system. Plenty of vendors will treat that as a checkbox to be added later, under pressure, when a corporate buyer's procurement questionnaire asks about it. Building the label in as a default is a different posture: it assumes the disclosure is part of the product rather than a compliance patch applied to it.
Worth being precise about what that does and does not achieve, because we make this point often: disclosure is not accuracy. Telling customers they are speaking with an AI does not make what the AI says true, and it does not settle who is answerable when the bot promises something the company does not offer. That question is answered by what the company wrote down as its standard, and by whether anyone checks the bot against it. Every operator who deploys a widget like this inherits that responsibility on day one, whoever built the tool.
What the company says it learned
Asked about a moment when doubt changed the product, the answer was about setup rather than models.
The configuration process had become too difficult for users. The convenient response would have been to write better documentation and conclude that customers needed more explanation. Instead, the company says, it treated the difficulty as a verdict on the product and rebuilt parts of the onboarding and configuration experience.
"A product is not only about whether the technology works," is how Teydra Lab puts the lesson. "It also has to be understandable and usable for real people."
That is not an original insight, but it is a rare one to act on. Rewriting onboarding is unglamorous work with no demo value, and the alternative, blaming the user, is always available and always cheaper.
Where it stands
The company is direct about the stage: the platform is under active development, with bugs to fix and features still to expand. It does not publish customer references, and we have not verified any live deployments, so nothing here should be read as a claim about who is running it today. What can be checked is the product itself, the published price, and the legal and compliance documents sitting behind it.
The stated five-year ambition is a combination that sounds simple and almost never survives contact with a roadmap: unusually strong control paired with genuinely simple use. Most products in this category pick one. They are either a toy with three settings or a platform that needs a consultant.
If it lands, the thing worth watching is not the chatbot. It is whether "pay once, control it yourself" turns out to be a viable way to sell AI software to small businesses at all, or whether the industry's subscription reflex proves stronger than the customer's preference. The product is at teydralab.com.
---
This is a Partner Story: a written interview with a company building with AI, submitted through Submit Your Story and published free of charge. Statements about the company's product, customers and results are its own. Building something with AI? Tell us about it.
