Go to the “Knowledge” tab. This is where you’ll set up everything and add your new embeddings. Make sure you have selected an environment in the top select, or the button will be blocked until you do so.

Manual

The simplest way to create a new embedding is to do it manually. Click the green “+” button or the “Create New” option under Build Knowledge, and a new modal will appear.

From there, set a title and fill in the content. The behavior and type fields will be locked to “context” and “manual”, since that’s the mode you’re currently using.

The Title is not part of the embedding itself. It’s only there to help you manage and organize your entries more easily. If you want the title to be included in the actual embedded data, you’ll need to also add it manually inside the Content field.

Embeddings based on Posts

AI Engine includes built-in tools to help you generate embeddings based on your WordPress posts. This includes automatic synchronization when posts are modified and automatic deletion of embeddings when posts are deleted.

You can also choose a specific post type to use for embedding sync. This gives you flexibility, you can create a custom post type (CPT) specifically dedicated to embeddings. This way, you can design advanced behaviors tailored to your content structure, using embeddings that are cleanly separated from your regular content. This setup is especially useful if you’re aiming to create dynamic interactions or content-aware chatbots.

You can also do that manually, without the sync, by using the “Push All” button and selecting a post type; this will go through all the posts from that post type and create embeddings based on their content. If the sync is not enabled, then it’s just like manual embeddings; if the sync is enabled, they will also auto-update.

Advanced Sync Behaviors

You can enable Expert Mode to get more control over how your posts sync. Within this mode, you can turn on “Rewrite Content”, which forces the embedding to be rewritten every time a sync is triggered, even if the post content itself hasn’t changed.

This is useful if some of your content comes from custom fields (like ACF) or third-party builders, since that data might not be part of the main post body. The Content Parser addon will update its written data when this option is enabled.

Embeddings based on PDF

To create embeddings from a PDF file, go to the Build Knowledge section on the right side of the Knowledge tab and click the “Upload PDF” button. You can then follow the step-by-step process to create embeddings.

First, make sure you understand what embeddings are and what they’re used for. Please read this documentation to get a clear overview of the concept.

1. Go to the “Dashboard” of AI Engine:

2. Enable the “Knowledge” option to display the “Knowledge” tab for later :

3. Go to Settings > Knowledge and make a new Environment with the type “OpenAI Vector Store“.

4. Let’s go to the OpenAI API platform where we will be able to create our new vector store.

5. Click the “Create” and configure based on your preferences.

6. Copy the “ID” and set it to AI Engine settings.

7. Select the Vector Store in your chatbot settings.

If you get this message, this is either because you did not select any environment for this vector store, or the environment is not the same as the chatbot you are trying to use right now. For this, go to the environment settings again and choose the right environment.

7.b. Select an OpenAI environment.

8a. Add files to your Vector Store. Vector Stores are entirely configurable on the OpenAI API platform, so if you want to modify it, add files, or remove files, there’s nothing to do in AI Engine after this point. Now, everything will happen on OpenAI, and your chatbot will query this vector store remotely.

8b. You can also use the Knowledge tab to add vectorized data as you would with a normal vector database like Pinecone. This includes syncing all of your posts automatically inside your vector store.

9. Ask about the files in your chatbot.

This documentation has not been written yet.

First, make sure you understand what embeddings are and what they’re used for. Please read this documentation to get a clear overview of the concept.

Go to the “Dashboard” of AI Engine:

Enable the “Knowledge” option to display the “Knowledge” tab:

You should now see a new tab called “Knowledge”. This is where you’ll manage all your embeddings within AI Engine.

From this tab, you can create, upload, organize, and maintain your knowledge bases.

Now you’ll need to create (if you haven’t already) and link your Chroma account and environments with AI Engine.

Click here to create your Chroma account and you’ll receive a $5 free gift from the AI Engine team. 🎉

Let’s create a new database for your embeddings. Simply click the “Create Database” button.

Once your database is created, you’ll be prompted to select a data source, but don’t worry, you’re already done for AI Engine. You can now go back to WordPress.

Before leaving Chroma, though, make sure to copy your API key as you’ll need it to connect AI Engine to Chroma.

Go back to the main dashboard by clicking your name in the top-left corner, or go directly to Chroma. Then navigate to Settings → API Keys → Create to generate your API key.

Go back to WordPress, then copy your API key and paste it in AI Engine.

To do that, go to Settings → Knowledge → Environments for Embeddings, make sure your environment type is set to Chroma, and then paste your API key in the corresponding field.

If everything is set up correctly, click the “Connect to Chroma Cloud” button. You should see a success message confirming the connection.

Here, we’re using Chroma Cloud, but Chroma also offers a self-hosted version if you prefer to manage your own data. You can find it on GitHub.

You’ll see a “Collection” named here, you can leave it as it is or delete it. Think of a collection as a compartment in your database. Check the Advanced section to learn more.

You’re done! 🎉

Now, go back to the Knowledge tab, select your environment and start adding your embeddings.

You can read this documentation to explore the different ways to create or import them, depending on your workflow and the type of data you want to use.

Don’t forget to select this environment in your chatbot settings to actually use it.

Advanced

Embedding Model

Chroma Cloud includes its own embedding model (Qwen3) to convert your content into vectors. You can leave this as is, or choose a different model if you prefer.

To change it, go to Environment Settings → Advanced → Embeddings Model, and select AI Engine as your embedding model.

It will now use the model you’ve selected in Settings → AI → Default Environments for AI → Embeddings.

Collections

With a single Chroma account and one database, you can create multiple Knowledge environments in AI Engine (using the same API key).

Each environment, referring to its own collection, can store its own separate data, allowing you to maintain unique knowledge bases for multiple chatbots. For example, different contexts or projects, all running at the same time from a single Chroma database.

Simply create a new environment with the same settings, but change the collection. You’ll now have two separate environments that can be used for different chatbots, each with its own unique knowledge base.

First, make sure you understand what embeddings are and what they’re used for. Please read this documentation to get a clear overview of the concept.

1. Go to the “Dashboard” of AI Engine:

2. Enable the “Knowledge” option to display the “Knowledge” tab for later :

3(a). Go on Pinecone’s website and access your dashboard by login in. There, create a new Project with the free tier starter plan.

3(b). Then you should be presented with your default API key — make sure to copy it, as you’ll need to register it in AI Engine to enable API-based functionality.

If you missed it, don’t worry. Head to the “API Keys” tab and you should see a “Default” one there. If you can’t see it, please create one using the button “Create API Key“.

Paste your API key somewhere safe, like in a note for now — we’ll first complete everything on the Pinecone side, and then register all the necessary details in AI Engine in one go.

3(c). You can now create an index. Choose a name and dimensions.

The dimensions should match what you have under ‘Default Environments for AI > Embeddings‘ inside the AI Engine settings. Simply use the ‘Setup by model‘ option and choose the one you want to use.

By default AI Engine uses 1536 (ada-3), so unless you have changed it, make sure it matches. In the example here we are using ada-3-small but feel free to use large or ada-2 if you want to, just make sure the dimensions are matching!

3(d). If everything is done correctly you should see something like this.

You can now get the “Host” value that we will use inside of AI Engine. Now you should have both your API key and your HOST, this is all we need to set things up in AI Engine.

4. Go back to AI Engine and in the “Settings” tab, click on the “Knowledge” section and insert the value of your previously copied API Key :

5. Enter you API key and the “Server” should match with your HOST :

In the above settings, you’re able to set the Min Score and Max Results. When a user sends a message in your chatbot, AI Engine transforms this message into an embedding and compares it with your existing embeddings to find the most relevant matches. But you already understand how embeddings work, right?

The “match” score represents the similarity between the query and an existing embedding, typically ranging from 0% to 100%. The Min Score setting allows you to define the minimum relevance required to consider an embedding a valid match. You may have multiple matches, and the system will consider up to the number defined in Max Results (e.g., 10).

Keep in mind that adding embeddings to the context will increase the total token usage. The larger or more numerous the embeddings, the more tokens are consumed. If some embeddings aren’t appearing in the context, it might be because your Max Content Length setting is too low relative to your Max Results setting. Make sure those values are balanced to avoid cutting off relevant matches.

6. Now, access the “Knowledge” tab of AI Engine:

7. On the top right of the table, you should see a dropdown to select your created env. If you don’t select an environment to work with here, the buttons will be disabled.

9. Everything should be setup correctly at this point. You can now add embeddings by clicking the “Add” button on the top left of this display.

10. Select your environment in your chatbot settings.

You’re done!🎉

Check this documentation to learn how you can add new embeddings.

If you’re not interested in knowing how it works and just want to get started, please at least read the Overview section to have a common understanding of what embeddings are.

Even though creating a knowledge base is a complex topic, AI Engine makes it simple. You can just set up your environment, sync your posts, and have your chatbot rely on your site content to generate responses.

In AI Engine, you can build your knowledge base using different services (ranked from easiest to most complex):

• Chroma
• OpenAI Vector Store
• Qdrant
• Pinecone

Then add your embeddings following this documentation here.

Quick Overview

When a user asks your chatbot a question, the chatbot first converts that question into an embedding. The system then searches your vector database to find the chunks of your documents whose embeddings are most similar (i.e., most relevant) to the question’s embedding. If a match is found, the matching embedding(s) content is added to the context.

This isn’t a keyword search, just having a keyword in your query doesn’t guarantee a match with an embedding.

You cannot ask the chatbot to “search the database” or “search the embeddings”—the chatbot doesn’t know that embeddings exist, nor can it directly access them. You also can’t ask for events happening “this weekend” unless the current date is explicitly provided in the same message. The embeddings match is based on similarity to the vectorized “this weekend” value, not the actual date it corresponds to.

Think of it more like a cloud of points in space, where each point represents a piece of embedded content. When a user submits a query, it’s converted into a vector (by an AI model specifically designed to do this) and placed in that space. The system then looks for the closest points (i.e., most semantically similar content) around that position.

So it’s about conceptual proximity, not exact word matching. Even if a keyword is present, it won’t match unless the meaning of the query is close to the embedded content.

If you are looking for a real search content system, you’ll need to build your own RAG solution using filters and/or function calling. You can learn more about this in the Dynamic Context section.

Learning about Knowledge

How does AI work?

At its core, AI is like a super-smart assistant that learns from data. Modern AI models, such as those based on large language models (LLMs) like GPT, work by processing huge amounts of text and patterns sourced from the internet or custom data. When you ask a question, the AI predicts the best response word by word, based on what it has “learned”.

Context is the “background story” that helps AI understand what you’re really asking. Without it, AI can give vague or incorrect responses. For example, if a customer asks, “How do I return an item?” on your e-commerce site, the AI needs context like your return policy, shipping details, and product specifics to respond helpfully.

In a business setting, context ensures the chatbot doesn’t hallucinate (make up facts) or provide outdated/generic info. By building a knowledge base, you’re giving the AI the right context from your own content, making interactions more personalized and trustworthy.

You can see in the example above a regular chatbot with no specific context, and on the right, the usage of embeddings and function calling to create a solid context. Here we’re talking about embeddings, but if you also use function calling you can create behaviors such as letting the AI model call a function to retrieve the purchase, check the license validity, then query the embeddings knowledge base to compare it with the return policy, and finally generate an educated response.

How does the AI use context?

To start, your AI is using instructions as its base guidelines. Technically, you could include all of your information there, what your business does, your contact information, how it should respond, your policies, and more. The AI could then use those instructions to answer.

But AIs have a limited context size, so for a business this isn’t something that could or should be done. For example, if you run an e-commerce site with hundreds or thousands of references, you can’t just input all that data and expect the AI to always pick the right piece of information when needed.

That’s where dynamic content comes into play.

This is exactly like a web search. When you look for information yourself, you don’t open all 1.5 million Wikipedia articles and read through them until you find what you need. Instead, you search for a specific topic related to what you’re looking for, and then you use that article, right? That’s exactly what we’re going to do with your chatbot.

Dynamic Context through RAG

RAG stands for Retrieval-Augmented Generation. It’s a smart way to make AI more accurate by combining two steps:

1. Retrieval: The AI searches your knowledge base for relevant information related to the user’s question.
2. Generation: It then uses that retrieved info, plus its built-in knowledge, to generate a natural-sounding response.

In simple terms, RAG is like having a librarian (retrieval) who finds the right wikipedia article, and then a storyteller (generation) who summarizes them into an easy-to-understand answer, where the “wikipedia” here would be your own knowledge base, your own corpus of data. This prevents the AI from relying only on its general training, which might not know your business details.

In AI Engine plugin, RAG helps your chatbot pull from your WordPress content or uploaded files to give spot-on replies. You can sync all of your posts, products, PDF files and more to build your knowledge base.

To make RAG work, we don’t rely on an ordinary database like the one used on your WordPress site. Instead, we rely on a specific kind of data storing technology, for which you’ll need to create a dedicated space.

What is a Vector Database and a Vector Store?

A vector database is a special type of database designed to handle “vectors” which are mathematical representations of data like text, images, or sounds. Think of a vector as a list of numbers that captures the essence or meaning of something. For example, the sentence “I love cats” might be turned into a vector like [0.2, 0.5, -0.1, …] based on its semantics.

This vectorization process is not performed by an algorithm or the plugin itself; it is done by AI, by a model that is not designed to generate text but to create these embeddings. The one that will be used in your case is the one set in Settings → AI → Default Environments for AI → Embeddings by default.

The idea here is that with these coordinates we can represent your database content as a cloud of points in space, where all data that share a similar meaning are clustered close to each other. This will be useful later for finding data related to your user’s message, since the message will also be converted into a vector and all the data nearby in the cluster will be fetched as context.

AI Engine will act as the bridge between this vectorized database and the actual textual data of your embeddings.

A vector store is similar, it’s essentially a storage system (often part of a vector database) where these vectors are kept and organized. Think of it as a warehouse for these numerical fingerprints, making it easy to quickly find similar items. This is usually used to store files directly, like PDFs, DOCX, TXT, images, and more, instead of relying only on textual data.

Regular databases (like SQL or NoSQL) are great for structured data, such as customer names, prices, or dates. They store info in tables or key-value pairs and search using exact matches or simple filters. For example, a regular database might find all products priced under $50 by checking numbers directly.

Vector databases, on the other hand, are built for unstructured or semantic data:

• Search Style: They use “similarity search” (e.g., “find things like this”) instead of exact matches. This is perfect for natural language, where meaning matters more than keywords.
• Data Type: They handle high-dimensional vectors (hundreds or thousands of numbers per item), not just text or numbers.
• Speed and Scale: Optimized for quick queries on massive datasets, using techniques like approximate nearest neighbors (ANN) to find close matches without checking everything.
• Use Case: Ideal for AI tasks like recommendations or chatbots, while regular databases are better for transactions or reports.

In short, regular databases are like a phone book (exact lookups), while vector databases are like a search engine (finding related ideas).

You can imagine how, in a regular database, creating a relation like “Jaguar car brand has a a feline mascot, which is similar to MeowApps’ Nyao cat mascot” would be a real headache to design. In contrast, with vectors in space, all of these topics naturally interconnect with each other.

How Chatbot Uses Vector Databases and Stores

When you build a knowledge base for your AI chatbot:

1. Your content (like wordpress posts or PDFs) is broken into small pieces.
2. Each piece is converted into a vector using an “embedding” model, which understands meaning.
3. These vectors are saved in the vector store/database.

When a user asks a question:

• The question is also turned into a vector.
• The AI searches the vector store for the most similar vectors (using math to measure “closeness” in meaning, not just keywords).
• It retrieves those relevant pieces and uses them in RAG to generate the answer.

This process is fast and efficient, allowing the AI to handle complex queries. For the AI Engine plugin, this means your chatbot can quickly recall info from your site without retraining the entire AI model.