Data Collection & Usage
Information We Collect
DiscMeet collects only the information required to provide AI-powered voice transcription services:
- Discord server and user IDs
- Voice channel audio data during transcription sessions
- Transcription results, session notes, and related meeting or session metadata
- Display names or labels we store with a session when available (for example, how speakers appear in notes)
- If you use Dungeons & Dragons features: campaigns, characters, places, and items you save in the dashboard for that server
- Basic usage statistics for reliability and performance
How We Use Your Data
Your data is used solely for:
- Providing voice transcription
- Generating summaries and meeting insights
- Ensuring service reliability and performance
DiscMeet does not sell, share, or use your data for advertising or model training.
Data Storage & Security
- Audio files are stored temporarily in DiscMeet's private Cloudflare R2 bucket.
- Files are encrypted at rest and in transit using industry-standard encryption.
- Audio files are deleted after transcription finishes.
- Transcript text and related fields we keep in our database (including AI-generated notes and speaker labels we store with a session) are encrypted at rest in the database.
Voice audio retention
We do not keep voice recordings as a permanent archive. Audio is used for processing, then removed according to our retention rules below.
No Third-Party AI Storage
- No third-party AI provider stores or retains audio or transcripts.
- Audio exists only in DiscMeet's encrypted R2 bucket during processing.
- All AI inference is ephemeral and does not persist content.
AI Processing & Open Source Models
DiscMeet uses open source models for transcription and analysis.
Open source models perform inference in real time with no data retention.
No Data Retention
- Open source models do not log or retain input audio or transcript text.
- Processing occurs entirely in memory and is deleted immediately after inference.
- No data is used to train any external models.
Complete Data Ownership
All data (audio, transcripts, insights) belongs exclusively to DiscMeet and its users.
No third-party AI service has any long-term access or retention.
Consent & recording
Important: Your responsibility as a server owner
Recordings start only when someone on your server uses DiscMeet's commands; the server owner is responsible for making sure recording is allowed and that participants understand what is being captured, in line with the laws that apply to you.
Optional opt-in mode: Server owners can turn on a setting so that only people who have opted in—based on rules the owner configures, such as specific Discord roles or an explicit allowlist—are included in transcription and in merged audio for a session. When that mode is off, the same consent and legal responsibilities still apply to everyone in the channel.
Server owners must ensure they have proper consent from participants and comply with applicable laws and regulations in their jurisdiction.
Your Rights & Control
You have the right to:
- Request access to your data
- Request deletion of your data
- Opt out of data collection
- Export your data in a portable format
Data Deletion Requests
Contact us via our Discord server.
We process deletion requests within 30 days.
Third-Party Services (Essential Only)
DiscMeet uses:
- Discord – authentication and bot functionality
- Open source models – ephemeral AI inference for normal meetings (no input/output retention)
- OpenAI – live API processing for Role Playing Games mode features only
- Cloudflare R2 – temporary encrypted audio storage (EU region)
Important:
Normal Meetings: DiscMeet uses open source models for transcription and language processing with zero content retention. We do not use OpenAI or ChatGPT APIs for normal meeting transcription.
Role Playing Games Mode: DiscMeet uses OpenAI live APIs exclusively for RPG-specific features and processing. This applies only when using RPG mode features.
Data Retention Policy
- Audio files: Removed after transcription processing (not kept as a long-term library of raw voice)
- Transcription text: Retained for 30 days unless deleted earlier (stored encrypted at rest in our database)
- User settings: Retained until account deletion
- RPG campaign data: Kept while you use the feature and until you delete it or close the account, as described in the dashboard
- Usage & billing records: We retain recording metadata such as duration and timestamps for as long as your account is active, and afterward only as needed to meet our billing, tax, and legal obligations. This is metadata only — never audio or transcript content.
- System logs: 90 days (operational metadata only, no audio or full transcript content)
Contact Information
For privacy-related questions or concerns:
Last updated: August 2026