Table Canon
Automated notes, searchable history, prep ideas for GMs
Start new thread

Table Canon - Automated notes, searchable history, prep ideas for GMs

by
•
Focus on the session, not note taking. Simplify your prep time with a searchable memory-bank of what happened at your table. Table Canon turns uploaded game-night audio into a recap, outline, entity dossiers, searchable index, and prep ideas for tabletop RPG game masters. Memory evolves each session. No sign up required to get started, 6 hours of audio upload free! Message me that you saw the launch on Product Hunt and I’ll credit your account with an extra 10 hours.

Add a comment

Replies

Best
Maker
📌
Table Canon is for GMs who are tired of being the unpaid secretary between sessions. You upload the recording from game night. It writes the recap, keeps the people, places, and open threads, and you can ask what the table already decided on an earlier night. Prep is a short list of catalysts from what is still unresolved, so the next session starts from pressure instead of a blank page. Whisper plus a chat is fine for one night. This is for the campaign: week 8 still knows what happened in week 2, without pasting the whole history back in. You can look before you upload anything. The sample campaign is several real nights, including questions that pull from an earlier session: Try it free as a guest. No credit card. Free is 6 hours of processed audio, lifetime, which is about one long night or two short ones. Session audio stays on your account and is NEVER used to train models:
Google Cloud Run
Google Cloud Run Hackathon
How did you use Cloud Run for this launch?
Maker
Table Canon’s launch backend is two Cloud Run services from one image. The Audio API is the public site: Firebase-authenticated requests, 2Gi, concurrency 5, and one warm instance so signup and upload do not cold-start. A game master uploads the recording straight to Cloud Storage with a signed URL. The API only registers the session and enqueues a Cloud Task. The Audio Worker does the long job: split the file, run transcription, then build the recap, World, and prep. It is not public. Cloud Tasks calls it with an OIDC token for the Runner. The worker is 4Gi, concurrency 1, timeout 3600 seconds, and scales to zero, because a session is one heavy job and there is no traffic between game nights. The queue caps concurrent dispatches at 5, so a launch spike queues uploads instead of stampeding the database. Both services use the same Cloud SQL Postgres instance. Cloud Build ships the image to the worker first, then the API.