What a WordPress Knowledge Base Chatbot Does
Most WordPress sites already hold the answers customers ask for. They are spread across a help centre, an FAQ page, a returns policy, a PDF brochure and a few blog posts. A knowledge base chatbot pulls those sources together and answers in plain language, so the visitor never has to know which page holds the answer.
Bitsa AI answers from your own approved AI knowledge base: the website pages, FAQs, documents and product catalogues you choose to train it on. It does not pull generic internet content into replies. Because it understands natural language, context and intent, visitors can ask the way they talk ("my order came damaged, what now?") instead of picking from a rigid menu of buttons.
It can run on OpenAI GPT models, among other large language models, and replies in the visitor's language across 50+ languages, including English, Hindi and Hinglish.
How Is It Different From a Knowledge Base Plugin or Site Search?
A knowledge base plugin organises articles into categories. Site search matches keywords. Both leave the visitor to read and work out the answer; the chatbot does that step for them.
| Question | KB plugin / site search | Knowledge base chatbot |
|---|---|---|
| How does the visitor ask? | Keywords, or browsing categories | A full question in their own words |
| What comes back? | A list of articles to open | A direct answer drawn from your content |
| Answer spread over two pages? | Visitor reads both and combines them | Combined in one reply |
| PDFs and documents | Often not searchable | Trained on directly |
| Hindi or Hinglish question | Usually no match on English articles | Replies in the visitor's language |
| Visitor still needs a human | Leaves for the contact page, or leaves | Shares name, phone and email in the chat |
You do not have to remove your knowledge base plugin. Keep publishing articles there for search engines and readers who like to browse, and let the chatbot answer from the same articles for visitors who just want the answer.
Bitsa AI Chatbot for WordPress at a Glance
- Price
- The plugin is free on WordPress.org. It connects to a Bitsa AI account, which runs the AI: every new account gets a 7-day free trial with no credit card, and paid plans start at ₹299/month. See plans.
- Setup
- No code. Install the plugin, train the bot in your Bitsa AI dashboard, add your domain, then paste your Client ID into the plugin settings.
- What it learns from
- Your website URLs, sitemaps, PDFs, text documents and FAQs. It answers from that content, not from the open internet.
- Languages
- 50+ languages, including English, Hindi and Hinglish. It replies in the language the visitor writes in.
- Lead capture
- Collects names, phone numbers, email addresses and requirements inside the chat and records them in your dashboard.
- WooCommerce
- Answers product questions, helps shoppers find items and adds products to the cart through WooCommerce's own Store API.
- Speed
- One loader of about 16 KB gzipped, loaded asynchronously in the footer. The chat window loads after the page has finished loading.
- Requirements
- WordPress 5.0 or later and PHP 7.4 or later. Tested up to WordPress 7.1.
How to Turn Your WordPress Content Into a Chatbot
The plugin is a lightweight connector. AI responses, knowledge-base training and lead management run on Bitsa AI's cloud, so your WordPress server does none of the heavy work. The full walkthrough with screenshots is in how to add a chatbot to WordPress; the short version:
Install the plugin
WP admin → Plugins → Add New, search "Bitsa AI Chatbot", click Install Now, then Activate. It needs WordPress 5.0+ and PHP 7.4+.
Create your Bitsa AI account
Sign up at bitsaai.com. Every new account gets a 7-day free trial with no credit card.
Train the knowledge base
In the dashboard, open Knowledge Base and add your help-centre URLs, sitemap, PDFs, text documents or FAQs. Training takes minutes.
Authorise your domain
Under Domain Access, add your domain without https://. This covers www and subdomains, and stops anyone else using your quota.
Copy the Client ID
Go to Branding → Embed code and copy your Client ID. You can set the theme, launcher icon, greeting and bubble position here too, without code.
Paste and save
In WordPress, open Bitsa AI Chatbot in the left menu, paste the Client ID (or the whole embed script; the plugin extracts the ID) and click Save Changes.
What Should You Feed It?
A chatbot can only be as good as the content behind it. These sources give the best answers for support and pre-sales questions. For a closer look at how page training works, see a chatbot trained on your website content.
Help articles and docs
Setup guides, how-tos and troubleshooting pages. Add the help-centre URLs or the sitemap.
FAQ pages
The fastest win: it is already written as question and answer. You can also add FAQs in the dashboard.
PDFs
Product manuals, brochures, course prospectuses, rate cards and warranty terms that live in PDFs nobody downloads.
Policy pages
Shipping, refunds, cancellations, appointments and payment terms. These drive a large share of repeat questions.
Service and product pages
What you offer, who it is for and what is included, so the bot can handle pre-sales questions as well as support.
Text documents
Answers your team gives on WhatsApp or email every day but has never published. Write them up and upload them.
Leave out: old pricing pages, discontinued products, draft articles, internal-only notes and two pages that say different things about the same policy. If a source is wrong, the bot will repeat it confidently.
How Do You Keep the Answers Accurate?
Answer quality comes from the content you train it on. Follow these habits and most wrong answers never happen. The idea behind this approach, retrieving from your own documents before answering, is explained in our guide to RAG chatbots for websites.
- One source of truth per topic. If the returns policy appears on three pages, make sure all three match, or train on only the official one.
- Put facts in text, not images. Prices, timings and delivery areas inside a banner image are invisible to training. Write them in the page or the FAQ.
- Be specific. "Refunds in 5–7 working days to the original payment method" gives a usable answer. "Refunds are processed promptly" does not.
- Update the source when something changes. When a policy or price changes, edit the page, then add the updated page or file to the Knowledge Base so the bot learns the new version.
- Test with real questions. Before launch, copy 20 questions from your support inbox and WhatsApp chats, including Hinglish ones, and ask them in the chat window. Every weak answer points to content you need to add or fix.
Install the free plugin, train it on your help centre and FAQ page, and test it with your own questions during the 7-day free trial.
Download Free PluginDeflecting Repetitive Support Questions
Most support teams answer the same questions every day: delivery times, login resets, what a plan includes. A knowledge base chatbot answers them at any hour, so your team spends its time on the questions that need a person.
List your top repeat questions
Go through a week of email and WhatsApp chats. Make sure each common question has a written answer in your sources.
Let the bot take the first reply
Visitors get the answer straight away, even outside office hours, in English, Hindi, Hinglish or another supported language.
Collect details when a human is needed
When a question needs your team, the chatbot can collect the visitor's name, phone, email and requirements inside the conversation. The lead appears in the Bitsa AI dashboard for follow-up.
Hide it where it does not belong
Developers can turn the widget off on chosen pages with the bitsaai_chatbot_load_widget filter, for example on checkout with is_checkout().
Common Mistakes When Building a Knowledge Base Chatbot
Training on the whole site without checking it
A full sitemap also pulls in old offers and outdated posts. Review it, or add help and policy URLs one by one.
Expecting it to know what you never wrote down
If the answer lives only in a staff member's head, the bot cannot give it. Treat each unanswered question as a content gap.
Skipping the domain step
If the chat bubble does not appear after setup, the usual cause is that the domain is not added under Domain Access. The second is a caching or optimisation plugin: clear its cache. The plugin works with Cloudflare Rocket Loader and uses WordPress's native async loading on WP 6.3+.
Forgetting the privacy policy
Copy the suggested text the plugin adds under Settings → Privacy → Policy Guide into your privacy policy.
Frequently Asked Questions
Quick answers before you add a knowledge base chatbot.
Yes. In the Bitsa AI dashboard's Knowledge Base you can train the bot on website URLs, sitemaps, PDFs, text documents and FAQs, so manuals, brochures and policy documents become answerable alongside your help-centre articles. Use PDFs with real text rather than scanned images.
Bitsa AI answers from your own approved knowledge base, the pages, FAQs, documents and catalogues you train it on, rather than generic internet content. That means answer quality depends on your sources: if they are clear, current and consistent, answers will be too. When a question exposes a gap, add the missing information and update the source.
No. Keep your knowledge base or docs plugin for browsing and search engines, and train the chatbot on those same article URLs or your sitemap. The Bitsa AI Chatbot plugin is a lightweight connector that loads one script of roughly 16 KB gzipped, asynchronously in the footer, so it does not block page rendering or interfere with your existing help-centre setup.
The Bitsa AI Chatbot plugin itself is free to download from WordPress.org. It connects your site to a Bitsa AI account, which runs the AI. Every new account includes a 7-day free trial with no credit card required, and paid plans start at ₹299 a month. Check the Bitsa AI pricing page for the current plans and what each one includes.
Setup needs no code. In WordPress, go to Plugins → Add New, search for Bitsa AI Chatbot, then install and activate it. Create an account at bitsaai.com, open Knowledge Base and train the bot on your website pages, sitemap, PDFs or FAQs. Add your domain under Domain Access, copy your Client ID from Branding → Embed code, and paste it into the plugin settings. Before launch, ask it the questions customers ask you most and check the answers. If the chat collects names, phone numbers or emails, update your privacy policy: the plugin adds suggested text under Settings → Privacy → Policy Guide.
Bitsa AI understands and replies in 50+ languages, including English, Hindi and Hinglish (Hindi written in Latin script). It detects the language a visitor writes in and answers in the same language. If many of your customers write in another language, such as Marathi, test a few real questions during the free trial. Leads captured in the chat are recorded in your Bitsa AI dashboard, and on WooCommerce stores cart actions go through your own Store API. For any other tool you need connected, ask the Bitsa AI team before you rely on it.
Why Use Bitsa AI for Your Knowledge Base Chatbot
Free plugin, light footprint
Free on WordPress.org (GPLv2, tested up to WordPress 7.1). One loader of about 16 KB gzipped, loaded async in the footer.
Your content, your domain
Answers come from the sources you approve, and the bot runs only on domains you authorise in Domain Access.
Branding without code
Change the theme, launcher icon, welcome greeting and bubble position in the dashboard. No code, no plugin update.
Priced in rupees
Start with a 7-day free trial, no credit card. Paid plans start at ₹299/month; compare them on the chatbot pricing page.
Put Your Help Content to Work
Your docs, FAQs and PDFs already hold the answers. Install the free plugin, train it on those sources and paste the Client ID. Visitors get direct answers, and questions that need your team arrive as leads.
Start 7-Day Free Trial