Skip to content

Knowledge Base

Upload your company documents and let AI use them to generate accurate, company-specific content.

Overview

The Knowledge Base connects your business documents to Vellocity's AI workflows. Using AWS Bedrock Knowledge Base with RAG (Retrieval-Augmented Generation), every piece of generated content can reference your actual product information, case studies, pricing, and technical documentation.

How It Works

graph LR
    A[Upload Documents] --> B[AI Indexes Content]
    B --> C[Generate Content]
    C --> D[AI Searches Your Docs]
    D --> E[Accurate, Contextual Output]
  1. Upload your company documents
  2. AI indexes the content using Amazon Nova embeddings
  3. When you generate content, the AI searches your documents
  4. Relevant information is incorporated into outputs
  5. You get content that's accurate and specific to your business

Key Features

Document Upload

Upload multiple file types:

Format Best For
PDF Product docs, case studies, whitepapers
TXT Specifications, notes, guidelines
CSV/Excel Pricing data, feature matrices, comparisons
Word Proposals, reports, documentation

Universal Availability

Knowledge Base works across all AI Workflow categories:

  • Content Studio — Blog posts reference your product details
  • Analysis Hub — Competitive analysis uses your positioning data
  • Technical Builder — Code docs reference your architecture
  • Media Studio — Presentations include accurate product info
  • Solution Chat — Chat responses use your company knowledge

Scope Control

Control who can access Knowledge Base data:

Scope Description
User Only you can access
Team Your team members can access
Brand All users under your brand
Company Organization-wide access

Product Isolation

Keep product-specific knowledge separate:

  • Each product can have its own Knowledge Base
  • Content generation uses only the relevant product's data
  • Cross-product knowledge sharing is configurable

AWS-Native Security

All Knowledge Base processing stays within AWS:

  • Amazon Nova Embeddings — Vector generation
  • S3 Storage — Encrypted document storage
  • Tenant Isolation — Your data is completely separate from others
  • FTR Compliant — Meets AWS Foundational Technical Review requirements

Getting Started

Step 1: Upload Documents

  1. Navigate to Settings > Knowledge Base
  2. Click Upload Documents
  3. Select files from your computer
  4. Wait for processing (typically 1-3 minutes per document)

Step 2: Verify Indexing

After upload, check that documents show status "Indexed":

Status Meaning
Processing Document being analyzed
Indexed Ready for AI to use
Failed Check file format and retry

Step 3: Use in Workflows

  1. Open any AI Workflow (Content Studio, Analysis Hub, etc.)
  2. Toggle Knowledge Base on in the workflow settings
  3. Generate content — AI automatically references your documents
  4. Review output for accuracy

Best Practices

Upload Key Documents First

Start with your most important documents: product overview, pricing, case studies, and competitive positioning.

Keep Documents Current

Update your Knowledge Base when product details, pricing, or positioning changes. Outdated documents lead to outdated content.

Review AI Outputs

While the Knowledge Base improves accuracy significantly, always review generated content for correctness before publishing.

Plan Availability

Knowledge Base is available on all plans:

Feature GTM Starter GTM Accelerate GTM Command
Document Upload 2 credits/query 2 credits/query 2 credits/query
All Workflow Integration Yes Yes Yes
Product Isolation Yes Yes Yes
Cross-Brand Sharing - - Configurable