CHATBOTS

Slack data-dictionary gap escalation to Coda steward queue

When the concierge can't find a definition for an asked-about column, it logs the gap to a Coda owner-review table and tells the asker a steward has been notified.

CategoryChatbots
Enginesim
Difficultyintermediate
Triggerevent
Steps5
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerSlack question about a column or metricSlack
  • ActionLook up asset description in Snowflake metadataSnowflakeSnowflake
  • LogicBranch on whether a usable definition exists
  • ActionAppend or increment gap row in Coda steward queueCodaCoda
  • OutputReply in Slack confirming gap logged and steward notifiedSlack

What it does

Closes the loop on missing documentation. When someone asks about a column the catalog can't explain, the bot doesn't just say 'I don't know' — it files the gap into a Coda table, routes it to the table's data steward, and acknowledges the asker.

When to use it

When your dictionary is incomplete and you want the questions people actually ask to drive what gets documented next. Use it to build a demand-ranked backlog of undocumented columns instead of guessing.

How it works

  1. 1A Slack question about a column or metric arrives.
  2. 2The bot looks up the asset in Snowflake metadata.
  3. 3A logic branch checks whether a usable description exists. If it does, it answers and stops.
  4. 4If the description is missing or empty, the bot resolves the steward from the schema's owner tag.
  5. 5It appends a row to a Coda documentation-gap table with the asset, the asker, and a timestamp, incrementing a hit count if the gap already exists.
  6. 6It replies in Slack that the gap is logged and the steward notified.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect SlackChannels, DMs, threads, mentions.
  2. 2
    Connect SnowflakeWarehouses, queries, shares.
  3. 3
    Connect CodaDocs, packs, automations.
  4. 4
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  5. 5
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  6. 6
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.