Table Canon: TTRPG | DnD campaign memory

Product Introduction

1. Definition: Table Canon is a web-based Software-as-a-Service (SaaS) application designed for tabletop role-playing game (TTRPG) game masters. It functions as an AI-powered campaign memory and session preparation tool, technically categorized under productivity and gaming utilities.
2. Core Value Proposition: It exists to eliminate the friction between playing a TTRPG session and preparing for the next one. Its primary value is automating post-session documentation by transforming raw session audio into structured, searchable campaign data, allowing GMs to focus on gameplay and creative prep rather than manual note-taking.

Main Features

1. AI-Powered Audio Processing & Transcription: The core technology ingests uploaded game session audio files. It utilizes automatic speech recognition (ASR) to generate a transcript, which is then processed by a large language model (LLM) to extract structured narrative data. This technical stack enables the conversion of unstructured conversation into organized content without manual input.
2. Automated Campaign Memory Bank: This feature automatically generates a searchable database of campaign entities from the processed audio. It identifies and catalogs key elements such as Non-Player Characters (NPCs), locations, plot points, items, and player decisions. This creates a living wiki of the campaign that evolves with each uploaded session, serving as a single source of truth.
3. Structured Session Outputs (Recap, Outline, Prep Desk): The platform doesn't just provide a transcript. It synthesizes the analyzed data into multiple, immediately usable formats for the GM: a narrative session recap, a bullet-point outline of key events, and a "prep desk" with actionable ideas and unresolved threads for the next session. This multi-format output caters to different GM planning styles.

Problems Solved

1. Pain Point: It directly addresses "GM burnout" and "prep fatigue" caused by the arduous, time-consuming task of post-session note consolidation and the mental load of remembering intricate campaign details across multiple sessions and players.
2. Target Audience: The primary user persona is the dedicated Game Master (GM) or Dungeon Master (DM) for long-form campaigns in systems like Dungeons & Dragons (D&D 5e), Pathfinder, or other narrative TTRPGs. Secondary users include organized play coordinators or groups who share campaign documentation duties.
3. Use Cases: Essential for GMs running weekly/bi-weekly campaigns who struggle with note consistency; for groups with irregular attendance where recaps are critical; for complex, homebrew campaigns with vast amounts of lore; and for GMs who want to analyze player engagement and plot trends over time.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros: Dramatically reduces administrative overhead between sessions. Creates a searchable, centralized campaign bible automatically. The free tier (6 hours + potential 10-hour bonus) offers substantial value for testing. The web-based platform requires no software installation.
   • Cons: Currently relies on post-session uploads; live recording is noted as "coming soon," which is a workflow gap. Audio quality and cross-talk at the physical table can impact transcription accuracy. The value is contingent on the AI's correct interpretation of context and names (e.g., distinguishing "Kane" the NPC from "Cain" the city).

2. Key Alternatives & Differentiation:

   • Manual Note-Taking (Notion, OneNote, Obsidian): While flexible, these require constant manual effort during or after play. Table Canon differentiates by being entirely automated, turning passive audio into active notes.
   • Live Transcription Tools (Otter.ai, Descript): These provide a raw transcript but lack TTRPG-specific structuring. Table Canon's key differentiation is its specialized LLM layer that understands campaign context, identifying game entities and generating game-master-specific outputs like prep ideas, which generic tools cannot do.
   • Dedicated Campaign Managers (World Anvil, LegendKeeper): These are powerful for top-down worldbuilding and note storage, but they are primarily manual input systems. Table Canon's clear differentiation is its bottom-up, audio-first capture mechanism that populates such a database automatically.

Frequently Asked Questions (FAQ)

1. How does Table Canon handle privacy and my campaign audio? Table Canon's privacy policy should be consulted, but typically, audio is processed securely for transcription and analysis. Users should confirm if data is used for model training and if they retain ownership of the output content.
2. What audio file formats and recording setups work best with Table Canon? For optimal transcription accuracy, upload clear, high-quality audio (e.g., WAV, MP3) from a dedicated microphone. Recordings with minimal background noise and distinct speaker voices (e.g., using a central condenser mic or individual lapel mics) will yield the best entity extraction and recap quality.
3. Can Table Canon integrate with Virtual Tabletop (VTT) platforms like Roll20 or Foundry? Currently, Table Canon operates as a standalone web application. Direct VTT integrations are not mentioned; workflow involves uploading audio exported from your VTT session or recorded separately. Future API access could be a potential development.
4. How accurate is the AI at recognizing fantasy names and specialized TTRPG terminology? Accuracy depends on the underlying ASR and LLM models. It may struggle with highly unique names initially. The system likely improves through context within a single campaign. A best practice is to review and correct key entity names in the dossier post-processing to train the system for future sessions.
5. What happens after the free audio hours are used up? The product operates on a credit-based system for audio processing. After the initial 6-hour free tier (and any promotional bonus), users will need to purchase additional hours or subscribe to a plan for continued use, as indicated by the "credit your account" language in the offer.