Knowledge Base

Overview

Knowledge Base stores the content AI Puffer can search before it generates an answer or a piece of content. Use it for support answers, product details, documentation, policies, posts, pages, WooCommerce products, uploaded documents, and other source text you want AI Puffer to use as context.

Providers

Use Local, OpenAI or Google managed stores, Pinecone, Qdrant, or Chroma.

Manage Vector Stores

Create, select, or delete vector targets.

Add Data

Add data and manage source records.

Settings

Configure chunking, batches, content rules, semantic search, and other settings.

Semantic Search

Publish and configure a frontend vector search form.

Troubleshooting

Fix missing targets, dimension errors, and empty results.

Providers

Provider	Target name	How it works
Local	Store	Source chunks and vectors stay in WordPress. Choose a compatible embedding provider for indexing and search.
Google	Store	Google manages file search through your direct Google connection.
OpenAI	Vector store	AI Puffer sends the source data to OpenAI Vector Stores. No separate embedding model is selected in AI Puffer for this target.
Pinecone	Index	AI Puffer creates embeddings with the model you choose, then stores the vectors in a Pinecone index.
Qdrant	Collection	AI Puffer creates embeddings with the model you choose, then stores the vectors in a Qdrant collection.
Chroma	Collection	AI Puffer creates embeddings with the model you choose, then stores the vectors in a Chroma collection.

For Local, Pinecone, Qdrant, and Chroma, the store dimension must match the embedding model’s output dimension. Use the same model and dimension for both indexing and querying. For example, if your Pinecone index, Qdrant collection, or Chroma collection is 3072 dimensions, use a 3072-dimension embedding model when adding data and when searching that data later.

If the dimension does not match, the vector provider can reject the data or return unusable search results.

Local

Local stores source chunks, metadata, and embedding vectors in your WordPress database. No Pinecone, Qdrant, or Chroma account is required. Your WordPress hosting stores the data; AI Puffer Cloud does not host it. Local still needs an embedding model for indexing and search. Use a Cloud embedding model or a compatible model from your own provider. The text model generating answers can be a different provider. To create a store:

Open AI Puffer > Knowledge Base > Stores.
Click Create store.
Select Local.
Enter the Name and Dimension.
Click Create store.
Use Add source in Data to choose this store and the embedding model before adding content.

Local store creation with name and dimension fields

Creating the store sets its dimension, not its embedding model. When adding or searching data, use the same model and dimension. Models with equal dimensions are not necessarily compatible. If you switch embedding models, re-index the content into a store configured for the new model. Local indexing and semantic search through a hosted embedding model use that provider’s billing or Cloud credits. Local storage itself does not consume Cloud storage credits, but uses your WordPress database and hosting resources.

OpenAI

OpenAI Vector Stores use your OpenAI account directly.

Go to AI Puffer > Settings > AI.
Select OpenAI as the AI provider.
Enter your OpenAI API key.
Sync models if needed.
Go to AI Puffer > Knowledge Base > Stores to create or refresh OpenAI vector stores.

OpenAI Vector Stores do not require a separate embedding model selection in Knowledge Base. OpenAI handles file storage, chunking, embedding, and vector search on its side.

Pinecone

Pinecone is configured from the Integrations settings.

Go to AI Puffer > Settings > Integrations.
Select Pinecone.
Enter your Pinecone API Key.
Click Sync Indexes to load indexes from Pinecone.
Go to AI Puffer > Knowledge Base > Stores to create, refresh, or delete indexes.

When you create a Pinecone index in AI Puffer, enter the dimension that matches the embedding model you plan to use.

Qdrant

Qdrant requires both an endpoint URL and an API key.

Go to AI Puffer > Settings > Integrations.
Select Qdrant.
Enter your Qdrant URL.
Enter your Qdrant API Key.
Click Sync Collections to load collections from Qdrant.
Go to AI Puffer > Knowledge Base > Stores to create, refresh, or delete collections.

When you create a Qdrant collection in AI Puffer, enter the dimension that matches the embedding model you plan to use.

Chroma

Chroma uses endpoint, tenant, and database settings.

Go to AI Puffer > Settings > Integrations.
Select Chroma.
Enter your Chroma URL. For Chroma Cloud, you can use https://api.trychroma.com
Enter your Chroma API Key if you use Chroma Cloud or an authenticated server.
Enter the Tenant.
Enter the Database.
Click Sync Collections to load collections from Chroma.
Go to AI Puffer > Knowledge Base > Stores to create, refresh, or delete collections.

For local Chroma, the default tenant is default_tenant and the default database is default_database.

Embedding Providers

Local, Pinecone, Qdrant, and Chroma store vectors that AI Puffer creates with a selected embedding model. Before adding data to these providers, configure the embedding provider you want to use in AI Puffer > Settings > AI. Supported embedding providers include AI Puffer Cloud, OpenAI, Google, Azure, OpenRouter, and Ollama where available. For Cloud, connect the account and refresh its models; you do not enter a provider API key. The selector shows actual model names. Supported output dimensions come from the model’s capabilities; a store is not limited to 512 dimensions. The selected embedding model must match the dimension of the Pinecone index, Qdrant collection, or Chroma collection. xAI is not an embedding provider or vector store provider in the current integration. xAI chatbots, forms, and text workflows can still use retrieved Knowledge Base context from Local, OpenAI, Google, Pinecone, Qdrant, or Chroma because AI Puffer sends that context as text.

Manage Vector Stores

Use AI Puffer > Knowledge Base > Stores to create, refresh, inspect, or delete vector targets. The Stores tab manages Local stores, OpenAI and Google managed stores, Pinecone indexes, Qdrant collections, and Chroma collections. The Data tab uses these targets when you add content.

OpenAI Vector Stores

Add your OpenAI API key in AI Puffer > Settings > AI.
Go to AI Puffer > Knowledge Base > Stores.
Select OpenAI as the provider.
Click Create Store.
Enter a store name.
Click Create.

OpenAI handles the vector store search on its side. AI Puffer stores a local source record so you can see what was added.

Pinecone Indexes

Add your Pinecone API key in AI Puffer > Settings > Integrations.
Go to AI Puffer > Knowledge Base > Stores.
Select Pinecone as the provider.
Select the embedding model you plan to use.
Click Create Store.
Enter an index name.
Enter the dimension for the selected embedding model.
Click Create.

Use the same embedding model when you add data to the index and when a module searches that index.

Qdrant Collections

Add your Qdrant URL and API key in AI Puffer > Settings > Integrations.
Go to AI Puffer > Knowledge Base > Stores.
Select Qdrant as the provider.
Select the embedding model you plan to use.
Click Create Store.
Enter a collection name.
Enter the dimension for the selected embedding model.
Click Create.

Use the same embedding model when you add data to the collection and when a module searches that collection.

Chroma Collections

Add your Chroma endpoint, tenant, database, and API key in AI Puffer > Settings > Integrations.
Go to AI Puffer > Knowledge Base > Stores.
Select Chroma as the provider.
Click Create Store.
Enter a collection name.
Click Create.

Chroma collections do not require a dimension when they are created in AI Puffer. Use the same embedding model when you add data to the collection and when a module searches that collection.

Use Refresh when you need AI Puffer to fetch the latest stores, indexes, or collections from the selected provider. To delete a target, use the available action in the Stores table.

Add Data

Use AI Puffer > Knowledge Base > Data to add new source data and manage existing source records. Before adding data:

Go to AI Puffer > Knowledge Base > Data.
Select a provider.
Select the target vector store, index, or collection.
Click Add source.
Choose the target store and, for Local, Pinecone, Qdrant, or Chroma, its embedding model.
Choose Website, Q&A, Text, or Files and add the content.

Q&A

Use Q&A for short answers that should be easy to retrieve later.

Select Q&A.
Enter the question.
Enter the answer.
Click Add Q&A.

AI Puffer stores the pair as text:

Q: question text
A: answer text

Text

Use Text for policies, instructions, product notes, support snippets, or any source text that does not already exist as WordPress content.

Select Text.
Paste the source text.
Click Add Text.

Files

Use Files when the source is already in a document.

Select Files.
Click Choose files.
Select one or more files.

Files start uploading and training after selection. Supported file extensions:

.pdf, .docx, .txt, .md, .csv, .json

For Local, Pinecone, Qdrant, and Chroma, AI Puffer extracts text, splits large files into chunks, creates embeddings, and stores each chunk in the selected index or collection. File chunks can be embedded in batches to reduce the number of embedding API requests. File size is limited by your WordPress/PHP upload settings. OpenAI Vector Store uploads also use OpenAI’s file limits.

Website

Use Website when the source is WordPress content.

Select Website.
Choose All or Choose items.
Select the content types.
If using Choose items, select the individual published items.
Click Add Items.

Website training uses published content. Posts and pages are selected by default. WooCommerce products appear when WooCommerce is active. Public custom post types can also appear. When WordPress content is indexed, AI Puffer builds the source text from the URL, title, excerpt, content, public custom fields, public taxonomies, and available WooCommerce product data.

Manage Data

The source table in the Data tab shows the local records created while adding data.

Column	What it shows
Status	Trained, Processing, Failed, or another provider status.
Item	Post title, text preview, file name, or source identifier. Provider, target, and embedding details appear below the item name.
Type	Site Content, Text, Q&A, File Upload, or User Upload.
Updated	Last update time and relative age.
Actions	Available actions for the source.

Available actions:

Action	Use it for
View	Review the stored source preview.
Edit	Edit a text source and save it again.
Retrain	Re-index a WordPress content source after the content changes.
Delete	Remove the source from the selected knowledge store and from the local source table.

Settings

Open AI Puffer > Knowledge Base > Settings to configure Knowledge Base behavior. Knowledge Base has Data, Stores, Search, and Settings tabs. The sections below describe the settings controls; some screenshots show their earlier tab layout.

Tab	What it contains
Chunking	Document chunking controls.
Batches	Embedding batch controls.
Content Rules	Rules for which WordPress content fields are embedded.
Others	Semantic Search and user upload visibility.

Chunking

Document chunking controls how AI Puffer splits large uploaded files and WordPress Website content before embedding them for Local, Pinecone, Qdrant, or Chroma.

Setting	Default	Range	Use it for
Avg chars per token	`4`	`2` to `4`	Estimates how many characters equal one token.
Max tokens per chunk	`3000`	`256` to `6000`	Sets the maximum chunk size before embedding.
Overlap tokens	`150`	`0` to `1000`	Repeats a small part of the previous chunk so context does not break sharply.

Use smaller chunks when an embedding provider rejects long input. Keep some overlap for long documents where meaning continues across sections. Some embedding models have lower hard limits than the visible maximum, so AI Puffer may apply a safer model-specific cap during indexing. OpenAI Vector Store file uploads use OpenAI File Search chunking instead of the Pinecone, Qdrant, and Chroma chunking settings above.

OpenAI File Search setting	Default	Range	Use it for
Indexing strategy	`Auto`	`Auto` or `Custom`	Lets OpenAI choose chunking automatically, or lets AI Puffer send custom chunking values.
Max chunk size tokens	`800`	`100` to `4096`	Sets the maximum OpenAI File Search chunk size when Custom is selected.
Chunk overlap tokens	`400`	`0` to half of max chunk size	Repeats part of the previous OpenAI File Search chunk when Custom is selected.

Batches

The Batches tab contains Embedding Batches, which controls how many file chunks AI Puffer sends to the embedding provider in one request.

Go to AI Puffer > Knowledge Base > Settings.
Open Batches.
Adjust the batch size for the embedding provider you use.
Wait for the settings autosave to finish.

If the Batches tab shows Upgrade, activate Pro before saving batch settings.

Provider	Default	Maximum
Google	`100`	`100`
OpenAI	`50`	`100`
OpenRouter	`50`	`100`
Azure	`50`	`100`
Ollama	`10`	`100`

For example, a batch size of 50 means AI Puffer sends up to 50 prepared file chunks to the embedding API at once. Larger batches can make file upload training much faster because they reduce repeated API calls.

Embedding batch settings apply to the supported providers’ chunked file uploads for Local, Pinecone, Qdrant, and Chroma. Cloud applies its own batch limits. Website training uses document chunking, but these batch-size settings do not change Q&A, Text, Website training, semantic search queries, or OpenAI Vector Store file uploads.

If a provider returns rate limit errors such as HTTP 429, lower that provider’s batch size and try again. AI Puffer can pause and retry file upload processing when the provider sends a retry delay, but lowering the batch size is usually better for accounts with stricter quotas.

Content Rules

Content Rules define which WordPress fields are included when Website training or list-screen indexing sends WordPress content to a vector target.

Go to AI Puffer > Knowledge Base > Settings.
Open Content Rules.
Click Configure.
Select a post type.
Adjust Basic Labels if you want different labels for source URL, title, excerpt, or content.
Enable or disable custom fields.
Enable or disable taxonomies.
For WooCommerce products, enable or disable product data such as SKU, price, stock, dimensions, and attributes.
Save.

If the Save button shows Upgrade, activate Pro before saving indexing rules.

Others

Open AI Puffer > Knowledge Base > Search for Semantic Search. Other Knowledge Base controls appear under Settings. The Others tab contains Semantic Search and admin visibility settings.

Semantic Search

Semantic Search publishes a search form that queries a Local store, Pinecone index, Qdrant collection, or Chroma collection from the frontend. Open the Search tab and configure the search target.

Select the storage provider: Local, Pinecone, Qdrant, or Chroma.
Select the index or collection.
Select the embedding model.
Set Number of Results.
Set No Results Text.
Test a query in Try semantic search.
Copy the shortcode.

[aipkit_semantic_search]

Semantic Search uses the global settings from this panel. It does not use OpenAI Vector Stores in the current UI. Use the same embedding model and dimension used when the Local, Pinecone, Qdrant, or Chroma data was added.

Admin Visibility

Setting	What it does
Hide user uploads	Hides chatbot upload records from the main Knowledge Base source table.

The post-list Index button is managed from AI Puffer > Settings > Utilities. See WordPress Utilities.

Control	What it does
Index Status column	Shows whether a post has already been indexed.
Index Status filter	Filters content by indexed or not indexed.
Add to Vector Store action	Sends selected posts to a vector target.

Troubleshooting

Provider target is missing

Configure the provider credentials, then sync or create the vector target again.

Pinecone, Qdrant, or Chroma training fails

Confirm the embedding model dimension matches the index or collection dimension.

Website content is missing fields

Check Knowledge Base > Settings > Content Rules for that post type.

WordPress list indexing controls are missing

Enable AI Puffer > Settings > Utilities > Index button and confirm the user role can access the vector content indexer module.

Semantic Search returns no results

Confirm the selected target contains trained data and the same embedding model is selected.

Overview

Getting Started

Setup

Modules

Utilities

Developers

Overview

Providers

Manage Vector Stores

Add Data

Settings

Semantic Search

Troubleshooting

Providers

Local

OpenAI

Pinecone

Qdrant

Chroma

Embedding Providers

Manage Vector Stores

OpenAI Vector Stores

Pinecone Indexes

Qdrant Collections

Chroma Collections

Add Data

Q&A

Text

Files

Website

Manage Data

Settings

Chunking

Batches

Content Rules

Others

Semantic Search

Admin Visibility

Troubleshooting

​Overview

Providers

Manage Vector Stores

Add Data

Settings

Semantic Search

Troubleshooting

​Providers

​Local

​OpenAI

​Pinecone

​Qdrant

​Chroma

​Embedding Providers

​Manage Vector Stores

​OpenAI Vector Stores

​Pinecone Indexes

​Qdrant Collections

​Chroma Collections

​Add Data

​Q&A

​Text

​Files

​Website

​Manage Data

​Settings

​Chunking

​Batches

​Content Rules

​Others

​Semantic Search

​Admin Visibility

​Troubleshooting

Overview

Providers

Local

OpenAI

Pinecone

Qdrant

Chroma

Embedding Providers

Manage Vector Stores

OpenAI Vector Stores

Pinecone Indexes

Qdrant Collections

Chroma Collections

Add Data

Q&A

Text

Files

Website

Manage Data

Settings

Chunking

Batches

Content Rules

Others

Semantic Search

Admin Visibility

Troubleshooting