The General Flow: From Data to Knowledge

KnowStack follows a clear, linear workflow. Understanding this flow helps you get value from the platform as quickly as possible. Every step builds on the one before it, so working through them in order gives you the best results.

The Four Stages

Think of KnowStack as a pipeline with four stages. Data flows in one direction: from raw sources, through AI processing, into a structured Knowledge Base that your team can search and use every day.

1. Collect — Bring your data into KnowStack (Data Collection page)
2. Generate — Let the AI turn that data into a structured Knowledge Base (Content Generation page)
3. Refine — Review, edit, and organize the generated content (Knowledge Base page)
4. Use — Search, share, export, and build on your knowledge (Knowledge Base page + Integrations)

Stage 1: Collect Your Data

Start on the Data Collection page. This is where you connect and import all the raw information that KnowStack will turn into knowledge. You can use as many sources as your plan allows, and you don't need to collect everything at once — you can always add more later.

• Crawl websites — Enter a URL and KnowStack extracts text from all pages on that domain. Great for product sites, help centers, and competitor pages.
• Connect email — Link your Gmail account via OAuth or any email provider via IMAP. Choose which labels to include, how far back to fetch, and which categories to exclude.
• Upload documents — Drag and drop PDFs, Word docs, text files, or CSVs (up to 25 MB each). The AI extracts the content automatically.
• Import meeting transcripts — Upload transcript files from tools like Otter.ai, Fireflies, or Fathom.
• Connect Telegram — Authenticate with your phone number and pull messages from private chats and group conversations.
• Connect Slack — Authorize your workspace via OAuth and choose which channels to sync.

Stage 2: Generate Your Knowledge Base

Once you have data collected, go to the Content Generation page. This is where the AI analyzes everything you've collected and produces a structured Knowledge Base.

1. Select which collected data to include — you'll see all your sources organized by type, with checkboxes to pick individual items. You don't have to include everything.
2. Configure your KB — give it a name, choose an output language, and customize which sections the AI should generate. You can toggle default sections on or off and add custom sections with your own titles and descriptions.
3. Click Generate — the AI processes your data and builds the KB in the background. You'll see a real-time progress screen showing which sections are being generated.
4. Get notified when it's done — generation typically takes 1–5 minutes depending on the amount of data. You'll receive a notification when it's ready.

Stage 3: Refine Your Knowledge Base

The AI gives you a strong starting point, but you should review and refine the result. Go to the Knowledge Base page to see your generated KB.

• Review the content — Read through each section to check accuracy, especially for critical business information like policies, pricing, and compliance details.
• Edit anything — Click any section to open the rich text editor. You can modify text, change formatting, add new content, or remove anything that isn't relevant.
• Reorganize — Drag and drop sections and subsections to reorder them. Add new sections or delete ones you don't need.
• Add content blocks — Each subsection supports paragraphs, bullet lists, numbered lists, headings, notes, warnings, code blocks, quotes, examples, summaries, and tables.
• Set up access control — Decide which team roles can view or edit the KB. Use the 'View As' feature to preview what different roles will see.

Stage 4: Use Your Knowledge

Now your team can get value from the Knowledge Base every day.

• AI Search — Ask questions in natural language (e.g., 'What is our return policy for enterprise accounts?') and get answers grounded in your KB content, with source references.
• Share with the team — Invite team members and assign roles so the right people have access. New hires can search the KB from day one.
• Export — Generate PDF exports (with company branding) or CSV exports for sharing with external stakeholders or offline use.
• Keep it current — Add new data sources at any time and use iterative generation to expand your KB without starting over. Manual edits are preserved through conflict detection.

The Ongoing Cycle

KnowStack is not a one-time setup. As your business evolves, the cycle repeats naturally:

1. New information arrives — a new product launch, a policy change, a new client onboarding
2. Collect the new data — crawl updated web pages, sync recent emails, upload new documents
3. Regenerate or expand — use iterative generation to add the new knowledge to your existing KB without losing your manual edits
4. Refine — review the new content and adjust as needed
5. Your team always has the latest information — no stale docs, no conflicting versions

Quick Reference: Where to Go for What