Task tier · Web

A chatbot that answers from your knowledge, not vibes.

On your site or WhatsApp, it answers visitor questions strictly from your knowledge base in Greek or English, captures leads, books or routes where it should, and hands cleanly to a human the moment it hits a limit, no hallucinated promises.

ModelFast model for speed, stronger model for hard questions
EvalMust pass 90%+ of our internal eval before go-liveHow we test
Cost per taskModel usage billed at cost, measured per task in your pilot
Connects toWebsite widgetWhatsApp BusinessMessengerHubSpotSlack
Autonomy
  1. Suggests
  2. Acts with approval
  3. Acts on its own

Answers from your approved content on its own and hands over to a person when unsure or when asked. How autonomy works

What does an AI website chatbot do?

A website / WhatsApp chatbot answers visitor and customer questions in real time, grounded in your knowledge base (RAG) so it speaks from your actual content, not guesses. It works on your site and on WhatsApp, in Greek and English, captures leads, can book or route, and hands off to a human with context when it reaches its limit. Task tier, running on Workers AI in your own Cloudflare account. It overlaps with a support agent but is lighter, front-of-site Q&A and capture rather than full ticket resolution. New to the distinction? See our guide on AI agent vs chatbot.

theagency47 · Updated June 2026
The spec sheet

How the Website / WhatsApp Chatbot actually works.

Job-to-be-doneAnswer visitor questions from your knowledge base, capture/route leads, hand off cleanly to a human
TierTask (bounded, high volume)
Underlying modelMix of LLMs: a fast model for speed, a stronger one for hard queries
TriggerVisitor opens chat (web) or messages your WhatsApp number
InputsKnowledge base (site, docs, FAQ), handoff rules, lead-capture fields, brand tone
Tools availableRAG over your KB, lead capture, calendar/booking, live handoff, CRM logging
Autonomous decisionsAnswer vs escalate, lead-capture timing, language detection, routing
Escalation rulesOutside KB → handoff, no guessing · User asks for human → handoff · Sensitive/complaint → priority routing
KPIs measuredConversations, containment %, lead-capture rate, handoff rate, CSAT, after-hours coverage
Eval suite20 test cases (grounding/no-hallucination, handoff triggers, bilingual answers).
What it does

The work it takes off your team.

Grounded answers

Answers strictly from your knowledge base, so it never invents prices, policies, or promises you can't keep.

Site + WhatsApp

Same brain on your website widget and your WhatsApp line: customers chat where they already are.

Bilingual

Detects and replies in Greek or English automatically, matching the visitor.

Lead capture

Qualifies and captures leads mid-conversation, writing them straight to your CRM.

Clean handoff

When it hits a limit it routes to a human with the full transcript: never a dead end or a loop.

24/7

Covers nights and weekends, so out-of-hours visitors still get answers and capture.

What it looks like in practice

A typical run.

  1. All day, Visitors ask questions on site and WhatsApp; the bot answers from your KB and captures leads.
  2. On limits, Anything outside the KB or requiring a human is routed with the transcript attached.
  3. After hours, Out-of-hours visitors still get grounded answers and are captured for follow-up.
  4. Ongoing, Unanswered questions feed back into the knowledge base to widen coverage.
FAQ

Questions about the Website / WhatsApp Chatbot.

Will it make things up?

No, it's grounded in your knowledge base with retrieval, and when a question falls outside that content it hands off rather than guessing. "No-hallucination" is an explicit eval category we test before launch.

Chatbot or full support agent?

The chatbot is lighter: front-of-site Q&A, capture, and routing. Orion, the support agent, resolves actual tickets end-to-end across a helpdesk. Many sites start with the chatbot and add the support agent as volume grows. See AI agent vs chatbot.

Does it need a knowledge base?

Yes, and that's the point. We build it on your content, and can pair it with a knowledge base agent that keeps that content current automatically.

How is it delivered?

As a Spark build, embedded on your site and WhatsApp, grounded in your content.

Free · 30 minutes · No deck, no pitch

Find the 3 processes eating most of your team's hours.

In a 30-minute discovery call we map how your team's week actually goes, name the three processes that cost the most hours, and tell you honestly whether an AI agent is the right fix. If it isn't, we'll say so.

We reply within one business day with a calendar link. Want it in writing? The AI Audit adds a written report with ROI projections, and the €297 is credited in full if you start a project within 30 days.

Prefer to write it down? Describe your agent in 2 minutes →

Part of