Skip to main content

Project Knowledge

Project Knowledge (or "Know-how") provides the specific context Harmony’s AI needs to analyze your conversations accurately.

By adding your company’s unique terminology, product details, and internal definitions, you ensure that the generated insights are factual and aligned with your business goals.

Why context matters​

Generic AI models may not understand industry-specific jargon or internal project names. Providing Project Knowledge bridges this gap, allowing the AI to:

  • Identify niche competitors accurately.
  • Understand complex pricing structures or discount rules.
  • Distinguish between similar-sounding internal products.
  • Correctly interpret "shorthand" used by your team in meetings.

If your Insights feel "generic," adding 2-3 paragraphs of specific Project Knowledge usually might help.

How to add Project Knowledge​

You can define knowledge for each project individually, ensuring the AI uses the right "brain" for the right initiative.

  1. Go to Insights and select your project.
  2. Navigate to the Configuration tab.
  3. Locate the Project Knowledge section and click Edit.
  4. Enter your specifics in the text field.
  5. Click Update to save.

What to include in your Know-how​

To get the most out of your Insights, consider including the following details:

  • Terminology and Jargon Define acronyms or technical terms used frequently by your customers or team. Example: "MQL refers to a Marketing Qualified Lead, which in our case must have a budget over $10k."

  • Product and Feature Names List your core products and specific features, especially those with common names that might be confused with general words. Example: "Our product 'Flow' is a workflow automation tool, not a reference to conversation flow."

  • Competitive Landscape List specific competitors and how customers typically refer to them. This helps the Competitor Analysis insight stay precise.

  • Pricing and Packaging Briefly explain your tiers (e.g., Solo, Plus, Max) and any "non-standard" deal structures your team is allowed to discuss.

How the AI uses this data​

When a meeting is processed, Harmony first reads your Project Knowledge. It then uses those definitions as a lens to view the transcript. When it answers a "Yes/No" or "Checklist" insight, it checks the transcript against your provided definitions to ensure the answer is objectively correct for your business.

Project Knowledge is private to your workspace. Harmony does not use this data to train public AI models.