Try for free
  1. Home
  2. /
  3. Blog
  4. /
  5. Tomás Ribeiro

Tomás Ribeiro

How chatbots work · pen name of the Vatdi editorial team

Explains retrieval-based chat in plain words: what the bot reads, why it can say "I don't know", how languages are detected, and where it still fails. Cites the research and the vendor documentation rather than borrowed statistics.

Flow of a store chatbot: question, retrieval from catalogue and policies, answer with product cards, grade, content fix
Chatbot fundamentals
159 views

Nine techniques that move a store chatbot from "answers something" to "answers correctly": grounding every reply in retrieved content, structuring the catalogue so attributes are findable, writing policy pages a retriever can use, steering with quick replies, handover rules, per-language testing, coupon conditions, guardrails on prices and promises, and a 30-question quality loop you re-run after every content change.

Read More
Store chat widget answering whether a backpack is waterproof, with a product card showing price and stock
Chatbot fundamentals
1,016 views

An AI chatbot for ecommerce answers shoppers from your own catalogue, policies and order data instead of from scripts. This guide explains how it differs from rule-based bots, live chat and generic AI chat, what it reads, what it does in a conversation, what it cannot do, how it installs on each platform, what it costs, and how to judge one in a ten-question trial.

Read More
Flow from website pages and PDFs through a one-line script to answers, leads and handover, with reports at the end
Chatbot fundamentals
479 views

An AI chatbot for your website answers visitors from the pages, documents and FAQs you already have, captures a lead when someone is not ready to act, and hands over to a person when a question needs one. This guide covers a services site, a SaaS site and a store, what to feed the bot, the one-line install, the three mistakes that waste it, and how to measure it.

Read More
Pipeline of a RAG chatbot: question, retrieval over your content, generation from the passages, answer or a decline
Chatbot fundamentals
617 views

A RAG chatbot answers in two steps: it retrieves the passages of your own content that match the question, then writes an answer from those passages only, declining when nothing matches. This guide walks one question through both steps, explains chunks, embeddings and keyword search in plain words, lists the ways retrieval still fails on a real store and the content fix for each, and shows how to test it.

Read More
Chat widget answering a French shipping question and a Portuguese follow-up in each visitor's language
Chatbot fundamentals
559 views

A multilingual AI chatbot does four separate things: detects the language of each message, retrieves the answer from content that may be written in another language, writes the reply in the visitor's language, and switches the widget's own labels. This guide explains how each step works, where it breaks (short messages, product names, mixed-language chats), what you configure, and how to test it before launch.

Read More

Ready to Add AI Chatbot to Your Store?

Join thousands of ecommerce stores using Vatdi for 24/7 customer support, lead capture, and sales automation.

Try for free Book a Demo
Free plan available No credit card required 5-minute setup