No coding skills needed to automate a price list or monitor sales — just explain it in plain language. Prosus, the Amsterdam-listed technology holding, has announced ToqanClaw, a platform that lets restaurant owners, shopkeepers and small entrepreneurs build apps, dashboards and automations simply by describing what they need, as if they were talking to a colleague. The goal is to put the power of artificial intelligence into the hands of the 5 million merchants who, according to the company, have been left behind by the digital revolution so far.

How ToqanClaw works: from natural language to digital tool

Users describe what they want — for instance, «I want a page where customers can book a table and see real-time availability» — and the platform turns that description into a working interface, complete with automation logic. This is not the industry’s first foray into low-code or no-code tools, but using LLMs to interpret intent and generate code on the fly promises to lower the barrier even further, removing the need to navigate complex configuration screens. Prosus hasn’t shared details about the underlying infrastructure, but it’s reasonable to assume ToqanClaw relies on large language models orchestrated through the cloud, with an abstraction layer that translates informal speech into application components.

Why merchants are ‘left behind’ by AI

While large companies invest in on-premise servers, GPUs and fine-tuning of proprietary models to analyze data and automate workflows, a restaurant or a small boutique struggles to find tailor-made tools. ToqanClaw is a direct response to this asymmetry, turning every merchant into a potential creator of digital solutions. Prosus sees a market that AI has overlooked until now: those without technical teams who handle daily operations that could benefit from a touch of automation. The conversational interface is the missing bridge.

The price of simplicity: control and data sovereignty

By their nature, tools like this run almost entirely in the cloud. Business data — revenue, customers, preferences — gets processed on third-party servers, with all the implications for privacy and GDPR compliance. For those operating in regulated industries or simply preferring to keep full control over their information, on-premise deployment remains a crucial alternative, though historically more complex and expensive. AI-RADAR dedicates a whole section to analyzing frameworks for on-premise LLMs, highlighting how self-hosted solutions are becoming more accessible even for mid-sized companies, albeit without the ‘describe and create’ convenience offered by platforms like ToqanClaw. The trade-off is sharp: immediate ease of use versus data sovereignty and transparency.

Horizons: conversational AI for the micro-entrepreneur

The direction is clear: as language models become more efficient and hardware barriers shrink, tools like this could eventually run on local devices, combining ease of use with data control. The growing focus on quantization and inference on consumer hardware suggests a future where a shopkeeper could manage everything locally, without sacrificing a natural-language interface. For now, ToqanClaw is an interesting test of how close AI can get to the everyday language of less digitized businesses, accelerating adoption that has been held back by technical complexity.