Twelve conversational agents across six industries. Voice and chat, in the languages people actually speak, for the ones who cannot type, cannot read English, or simply will not call. Built by a qualified dentist, BDS.
Agents that speak the language the caller opens with, say only what they can prove, and hand over to a person the moment a question needs one.
Most software assumes a person who types, reads English, and is comfortable with forms. A very large number of people are not that person. They are not a small edge case at the bottom of a spreadsheet; in India they are the majority, and every interface built without them quietly writes them out.
Take one, or the whole arc. Every engagement is led directly, with specialist collaborators brought in per build.
How the agent opens, what it will never say, and language that matches how the people you are trying to reach actually talk.
Whether to build or buy, where the agent should stop and a person should take over, and honest review against real conversations rather than a demo script.
Shipping the agent, wiring conversations into the systems already in use, and tuning from live transcripts.
Twelve conversational agents. Six industries. Two live in market. One cut lead qualification time by 35 percent, another automated 40 percent of inbound counselling calls.
That work was conversation design delivered on a client platform. The craft transfers; the client relationships do not. What it bought was pattern recognition: the caller who switches language mid-sentence, the person who abandons at the first form field, the moment a script stops sounding like a person. You do not learn that from a framework. You learn it from twelve of them going wrong in different ways. There is no paying healthcare customer.
The founder is a qualified dentist, BDS, who stopped practising but never stopped noticing what breaks between an organisation and the people it is meant to reach. The dental booking agent is our own proving ground, built and running rather than described.
Now working on the hardest version of that problem: an elderly person in India who owns no smartphone and does not read English.
See the evidence in full →If the people you are trying to serve cannot type, cannot read the interface, or never answer the message you sent, that is a conversation problem before it is a product one. Start with a written conversation. No deck, no demo call.
Start a conversation →