Anyone who has sold for years knows the pattern: the same follow-up questions, the same scheduling, the same chasing — every single day. And they know deals rarely fail on the offer. They fail on who answered first and who stayed with it.
The idea was to represent exactly those recurring sequences. But not generically. A toolkit that looks the same for every customer does not solve the problem — the process in an insurance agency is different from a car dealership or a public authority. And it has to stay close to human, or the customer notices by the second sentence that they are writing to a machine.
The channel was obvious: WhatsApp. Not because it is the newest tool, but because it is the number one communication medium. We expect businesses to use it increasingly over the coming years for customer care, support and sales. That is exactly what alkaio was built for — and is still being built for.
That there are two of us is not a coincidence. One half is the practice: knowing what a customer and what a salesperson actually want in a conversation, because you sat in those conversations for years. The other is the engineering: text and image processing, agent logic, the rules for when a model may answer and when it had better not. Without both it would be either a sales concept without a product or a model without an application.