Skip to main content

​
What It Does

  • Summarises text using AI at row or group level
  • Grouped summarisation support using custom key fields
  • Custom instructions to guide tone, focus, or summary length
  • Auto-resolves column conflicts (e.g., SUMMARY_1 if SUMMARY exists)

​
🏁 Getting Started

Text Summariser node configuration screenshot
1

Add the Node

Drag the Text Summariser node into your workflow.
2

Choose Columns to Summarise

Select one or more text columns—like @feedback, @notes, or @call_transcript.
3

(Optional) Enable Grouped Summarisation

Toggle Summarise Multiple Rows to summarise grouped data.
4

(Optional) Set Grouping Keys

Select columns like @account_id, @session_id, or @date to group by.
5

(Optional) Add a Custom Instruction

Tailor the summary (e.g., “Summarise for the sales team” or “List key takeaways”).
6

Run the Node

You’ll get clean, concise summaries per row—or per group.

​
Inputs

​
🛠️ Required Fields

  • Columns to Summarise (✅)
    One or more text fields to include in the summary.
    Why it matters: These columns will be merged (if multiple) and sent to the AI.

​
🎯 Optional Fields

  • Summarise Multiple Rows (⚪️)
    Toggle to enable grouped summarisation.
  • Grouping Keys (⚪️)
    Used when summarising multiple rows.
    Example: @account_id, @meeting_id
  • Summarisation Instruction (⚪️)
    Optional guidance like “Summarise in 2–3 bullets for execs” or “List issues and action items”.
  • Output Column Name (⚪️)
    Default: SUMMARY
    Choose a custom name like call_digest or feedback_snippet if preferred.
If a column named SUMMARY already exists, a suffix like SUMMARY_1 or SUMMARY_2 will be added automatically to prevent overwriting.

​
Output

Depending on configuration:
  • Row-by-row mode
    • All original columns preserved
    • Adds one summary column (auto-renamed if needed)
  • Grouped mode
    • Only grouping_keys + SUMMARY (or renamed) returned
    • One summary per group
Example output showing row-wise and grouped summaries

​
How It Works

  1. Validates your input selections
  2. Merges text if multiple columns selected
  3. Groups rows (if enabled)
  4. Applies your custom instruction (or default)
  5. Sends data to the AI for summarisation
  6. Outputs new column(s) with summaries

​
🚀 Example Use Cases & Prompts

Use CaseSetup Example
Feedback DigestSummarise @feedback per row
Meeting RecapGroup by @meeting_id, summarise @notes
Conversation ThreadGroup by @session_id, summarise @chat_log
News SummarySummarise each @article_text
Report TL;DRMerge @exec_summary + @analysis and return one summary

​
✨ Pro Tips

Use Grouped Summarisation for threads, sessions, or batched notes—it condenses 100s of rows into a clean, readable recap.
Tailor your summary with a prompt like “Summarise for support handoff” or “List top 3 themes.”
Rename your output to something specific—like call_summary or digest—to stay organized downstream.

​
⚠️ Important Considerations

Grouped summarisation changes the output shape — only grouping_keys + summary column are returned.
Larger text blocks take longer — long grouped transcripts or merged fields may increase processing time.

​
🛠 Troubleshooting & Gotchas

SymptomLikely CauseQuick Fix
No summaries returnedColumn not found or emptyDouble-check your input selections
Missing group fieldsGrouped mode is enabledOnly grouping keys + summary are returned
SUMMARY column missingName collisionLook for SUMMARY_1, SUMMARY_2, etc.

​
📝 FAQ

Yes — the selected fields are merged before summarising.
Absolutely — enable Summarise Multiple Rows and set your @grouping_keys.
The output column will be auto-renamed to avoid conflicts.

​
💰 Pricing

Each summary (row or group) consumes 2 credits.
Credits are charged per summary generated—not per row processed.
Example: 100 row summaries = 200 credits. 10 group summaries = 20 credits.

Turn long-form inputs into fast, usable insights. Great for call notes, feedback, and content distillation. 📝⚡