Skip to main content
Time: 15 minutes | Level: Advanced
Prerequisites: Knowledge Base Fundamentals

​
What You’ll Learn

  • Uploading multiple related documents
  • Cross-document querying
  • Handling conflicting data
  • Building comprehensive lookup systems
  • Use cases for multi-source knowledge

​
Why Multiple Sources?

Real businesses have data spread across multiple documents:
Document TypeContains
Product catalogNames, descriptions, SKUs
Price listCurrent pricing, tiers
Shipping ratesDelivery costs, zones
Inventory sheetStock levels, availability
Supplier catalogWholesale costs, lead times
Multi-source lookups let you query all of these simultaneously.

​
Setting Up Multiple Documents

1

Upload Related Documents

Add each document to your Knowledge Base:
  • Product Catalog.pdf - item descriptions
  • Price List Q1 2024.xlsx - current pricing
  • Shipping Zones.pdf - delivery rates
Each processes independently.
2

Verify Extractions

Check that each document extracted correctly:
  • Product names should be consistent across documents
  • SKUs or IDs should match where applicable
  • Dates should be recognized for version tracking
3

Note Common Identifiers

Identify how documents relate:
  • Product names (must match exactly)
  • SKU/Item codes
  • Category names
These act as “keys” linking data across sources.

​
Cross-Document Queries

Ask questions that span multiple documents:
Pull data from multiple sources in one query:
"What's the price, description, and shipping cost 
 for Product X to Zone 3?"
Peanuts searches:
  • Product catalog → description
  • Price list → unit price
  • Shipping zones → delivery cost
Returns a unified answer.

​
Handling Data Conflicts

When documents contain overlapping or conflicting data:
Scenario: Multiple price lists with different datesSolution: Peanuts uses the most recent document by default. You can specify:
"What was the price in the Q3 2023 price list?"
Explicitly references the older version.
Scenario: Same product name, different catalogsSolution: Specify the source:
"What's the price from the Wholesale Catalog 
 (not the Retail Price List)?"
Scenario: Two documents show different pricesBehavior: Peanuts flags the conflict:
“I found two prices for Widget X:
  • Retail Price List: $49.99
  • Promotional Flyer: $39.99
Which should I use?”
Scenario: One source should always take prioritySolution: In Knowledge settings, set document priority order. Higher priority sources override lower ones.

​
Linking to Helpers

Connect multiple knowledge sources to a single helper:
1

Open Helper Settings

Go to your helper → Settings → Knowledge.
2

Select Multiple Documents

Check all relevant documents:
  • ☑ Product Catalog
  • ☑ Current Price List
  • ☑ Shipping Rates
  • ☑ Inventory Status
3

Set Priority (Optional)

Drag to reorder if certain sources should take precedence.
4

Test Combined Queries

In the helper, ask questions that require multiple sources:
"Create a quote for 100 units of Item SKU-4521 
 shipped to Zone 2"

​
Use Case: Complete Order System

Build a helper that uses 4+ knowledge sources: Query flow:
  1. Customer asks for a quote
  2. Helper checks product exists (catalog)
  3. Looks up current price (price list)
  4. Verifies availability (inventory)
  5. Calculates shipping (shipping rates)
  6. Returns complete quote with all details

​
Organizing Large Knowledge Bases

By Category

Group related documents:
  • Products/
  • Pricing/
  • Logistics/
  • Policies/

By Version

Track document versions:
  • Price List v1 (Jan 2024)
  • Price List v2 (Apr 2024)
  • Price List v3 (Jul 2024)

By Client

Separate client-specific data:
  • Client A Rates
  • Client B Rates
  • Standard Rates

By Region

Geographic separation:
  • US Pricing
  • EU Pricing
  • APAC Pricing

​
Best Practices

Naming Conventions: Use consistent, descriptive file names that include:
  • Document type
  • Version or date
  • Scope (region, client, etc.)
Example: Retail-Price-List_Q1-2024_US.pdf
Avoid: Uploading dozens of tiny documents. Combine related data when possible—too many sources increases query time and potential for conflicts.

​
Exercise

Practice: Build a Multi-Source Quote System

Create a system using at least 3 knowledge sources:
  1. Create 3 sample documents:
    • Product list (5+ items with names/descriptions)
    • Price list (matching products with prices)
    • Shipping rates (3+ zones with costs)
  2. Upload all three to Knowledge Base
  3. Create a “Quote Generator” helper linked to all sources
  4. Test these queries:
    • “What’s the total for [product] shipped to [zone]?”
    • “Which products are under $50?”
    • “What’s the cheapest shipping option for [product]?”
  5. Intentionally create a conflict (same product, different price in two docs) and see how it’s handled
Success: All queries return accurate, combined data from multiple sources.

​
Key Takeaways

Remember: Multi-source lookups are like having a smart assistant who knows where everything is filed. Upload your documents, maintain consistent naming, and let Peanuts connect the dots across all your data.

​
Next Steps

Scheduled Reports

Automate insights from your knowledge

Webhooks & Automation

Trigger actions when knowledge is queried