AI Knowledge Base

AI Knowledge Base Turn Files into Scoped Knowledge

Upload PDFs, Excel files, web pages, and text into scopes. MonoChat chunks, embeds, and stores the data in your selected VectorDB — ready to power RAG pipelines and AI agent workflows.

Get Started For Free Chat on WhatsApp

Or start with $150 of credit, enough for a first month of Growth

Nova Store Assistant Answers from your documents

What is your return policy for online orders?

10:21

Source

returns-policy.pdf

Scope: Customer support

You can return online orders within 14 days of delivery, unused and in the original packaging.

10:21

Does the warranty cover the battery too?

10:22

Source

warranty-guide.pdf

Scope: Customer support

Yes. The warranty guide says the battery is covered for 12 months from the purchase date.

10:22

AI Knowledge Base Features

Build scoped knowledge from unstructured sources and use it across RAG pipelines and AI agents.

  • Multi-Source Ingestion

    Load unstructured content from files, URLs, and raw text. Keep everything centralized and searchable.

    • PDF, Excel, text, and more
    • Web page import
    • Bulk upload support
    • Metadata & tagging

  • Scoped Knowledge Domains

    Organize your data into scopes so every AI flow or agent retrieves from the right knowledge source.

    • Separate knowledge by topic
    • Per-scope access control
    • Assign scopes to agents
    • Avoid cross-domain leakage

  • Embedding Pipeline Controls

    MonoChat offers ready-to-use defaults, and also lets you control how embeddings are created.

    • Choose embedding model
    • Chunk size & overlap settings
    • Preprocessing rules
    • Optional advanced tuning

Core Components

Everything you need to ingest, embed, store, and retrieve knowledge reliably.

  • Smart Chunking

    Split documents into clean, retrieval-friendly chunks using defaults — or configure your own strategy.

  • Embedding Settings

    Pick the embedding model and configure how data is vectorized for fast semantic retrieval.

  • VectorDB Destination

    Store vectors in MonoChat embedded VectorDBs or connect your own VectorDB by adding credentials.

  • Governance & Lifecycle

    Update scopes, re-index data, and manage knowledge changes safely as your content evolves.

  • RAG-Ready Outputs

    Use scopes directly in your AI orchestrations so agents retrieve the right context at runtime.

  • Secure by Design

    Keep data isolated per workspace/scope and store embeddings where your compliance requires.

Practical Applications

Typical ways teams use AI Knowledge Base with scopes and RAG.

  • Customer Support Knowledge

    Create a scope for help center PDFs and policies so agents answer accurately and consistently.

  • Sales Enablement

    Embed catalogs, pricing sheets, and brochures into a sales scope for instant product answers.

  • Internal Operations

    Build SOP and onboarding scopes so teams and internal assistants work from a single source of truth.

Key Benefits

Core advantages you gain with scoped knowledge and configurable embeddings.

  • Higher Answer Quality

    Ground responses in your own data to reduce hallucinations and improve trust.

  • Full Control When Needed

    Use sane defaults or fully customize the chunking + embedding pipeline.

  • No Vendor Lock-in

    Use embedded VectorDBs or bring your own VectorDB and keep data where you want.

  • Composable with Agents

    Scopes plug into RAG, workflows, and AI tools for end-to-end orchestrations.

FAQ

AI knowledge base: common questions

Quick answers to the most common questions.

How do I train an AI chatbot on my own documents?

Create a scope in the MonoChat AI Knowledge Base and upload your PDFs, Excel files and text, or import web pages. MonoChat splits the content into chunks, creates embeddings and stores them in a vector database. Assign the scope to your AI assistant and it answers from that content.

Which file types and sources can I add?

You can add PDF and Excel files, plain text and web pages, with bulk upload for larger sets. Metadata and tags help keep the content organised and searchable.

What is a knowledge scope?

A scope is a separate knowledge domain, for example customer support, sales or internal procedures. Each scope has its own access control and is assigned to the agents and flows that should use it, so answers do not mix content from unrelated topics.

Do I need my own vector database?

No. MonoChat includes embedded vector databases that work with default settings. If you prefer to keep embeddings in your own infrastructure, you can connect your own VectorDB by adding its credentials.

How do I keep the chatbot answers up to date?

When your content changes, update the scope and re-index it. The assistant then retrieves from the new version, without rebuilding your flows.

Keep exploring

First month of Growth on us

Build Your Knowledge Layer with MonoChat

Upload documents, create scopes, and power your AI agents with reliable knowledge.

Start with $150 credit Chat on WhatsApp

One code per business • 30 days to redeem • No card needed

$150 to get started.

Claim