SKY Practice Follow-Up
Ananda — a warm check-in agent for people who recently completed a breath-and-meditation programme. It asks how the practice is actually going, listens without judgement if it has lapsed, answers common questions from the organisation's own knowledge base, and hands anything that needs a teacher to a human. A relationship call, not a survey and not a sales call.
Talk to the agent in your browser
Click “Talk to the agent” — Ananda will greet you in English and ask how your practice has been going. Try: “I stopped about two months ago.” · “What time of day is best?” · “I've forgotten part of the technique.” · “Can I speak to a teacher?” No phone number needed.
Allow microphone access when prompted. Ananda introduces itself as an AI assistant in its first sentence, answers truthfully if you ask whether it is AI, and will never describe or teach the technique itself — that always goes to a certified teacher.
Browser must support WebRTC + microphone (Chrome / Safari / Firefox). No calls are placed to any real phone number.
What it is doing underneath
Grounded in one knowledge base
Every factual answer comes from the organisation's own document — twenty pre-written responses sit in the prompt for instant recall, the rest is retrieved on demand. Nothing is answered from the model's general knowledge.
Escalates instead of guessing
When retrieval has no confident match, the agent says so plainly and files a follow-up for a human rather than inventing an answer. Forgotten technique, medical questions and complaints always route to a person.
Reads where the participant actually is
Practising, practising irregularly, lapsed, needing support, or wanting to go deeper — inferred from ordinary conversation, never asked as a questionnaire.
Structured outcome per call
Practice status, questions asked, issues raised, next best action and a short summary a team member can read instead of the recording — plus any follow-up ticket.
On the voice, and on being honest about it
This demo uses a synthetic voice. The agent states that it is an AI assistant in its opening sentence, answers truthfully whenever anyone asks whether they are talking to a person or to a specific teacher, and never claims to be one. It does not diagnose, does not promise health outcomes, and does not pressure anyone to practise or to enrol in anything.