Press a hotkey anywhere on your system, speak naturally, and Telvr transcribes your voice in real-time using Whisper. The finished text is automatically inserted at your cursor position — no copy-paste, no app switching.

Which languages are supported?

Telvr supports 50+ languages via OpenAI's Whisper large-v3 model. Language detection is automatic — just speak in your preferred language and Telvr handles the rest.

Do I need a subscription?

No. Telvr uses a pay-as-you-go model: EUR 3 per month infrastructure fee plus EUR 0.03 per minute of usage. No lock-in, no auto-renewal. You top up your balance and use it at your own pace.

Does Telvr work offline?

Currently, Telvr requires an internet connection for cloud-based transcription via Groq. A Community Edition with local processing using your own API key is planned for the future.

Which apps are supported?

Telvr works system-wide — it inserts text at the cursor position in any application. Email clients, chat apps, code editors, browsers, word processors — if you can type in it, Telvr works there.

All data is transmitted via TLS encryption. Audio recordings are not permanently stored after transcription. Groq processes your audio under a Data Processing Agreement (DPA). We do not sell or share your data.

← Blog2026-02-21

Automatic Meeting Notes: Never Miss a Detail Again

The Meeting Notes Problem

Every meeting produces information that needs to be captured: decisions made, action items assigned, context established. In practice, most of that information lives in someone's hasty notes, someone else's memory, or nowhere at all.

The tools that exist for meeting documentation tend to fall into one of two categories. The first is manual note-taking, which requires a dedicated note-taker who is barely participating in the meeting. The second is meeting recording tools like Otter.ai, Fireflies, and Fathom, which transcribe an entire meeting and require you to process a transcript after the fact.

There is a third approach that is faster than both: summarize immediately after the meeting using voice dictation with AI structure.

The Voice Summary Approach

The workflow is simple:

Have your meeting normally. Stay present — no note-taking.
Immediately after the meeting ends (within 1-2 minutes, while memory is fresh), open any text editor or document.
Hold your push-to-talk hotkey, switch to Meeting Notes mode, and speak your summary.
Release the hotkey. Within two seconds, a structured document appears.

What you speak sounds like: "We met with the product team to review the Q2 roadmap. Main decision was to delay the mobile app by six weeks. Sarah owns the timeline revision, due by end of next week. We agreed to cut the analytics dashboard from Q2 and move it to Q3. Next meeting is March 15th."

What you get is a structured document with sections for attendees, decisions, action items, and next steps — ready to share.

What Meeting Notes Mode Does

Telvr's Meeting Notes enrichment mode transforms spoken summaries into structured documents. The AI:

Identifies decisions and formats them as clear statements
Extracts action items and attributes them to named individuals with deadlines if mentioned
Structures the content with standard meeting note sections (context, discussion, decisions, actions, next steps)
Removes disfluencies from informal speech
Infers completeness — if you mentioned a decision without explicitly stating it as one, the AI recognizes the pattern

The output is immediately shareable without editing.

Example Output

Spoken input (60 seconds): "Quick product sync today. We went through the feature backlog for the month. Decided to prioritize the notification system because three enterprise customers complained last week. Tom said his team can deliver it in two sprints. Also talked about the onboarding flow — there's a drop-off at step three that nobody has investigated yet. Lisa is going to pull the analytics and share a report by Friday. The new hire starts Monday so Tom needs to set up the dev environment before then. Next sync in two weeks."

Generated output:

Meeting: Product Sync

Decisions

Notification system prioritized for current sprint cycle, based on enterprise customer feedback

Onboarding flow drop-off at step 3 to be investigated

Action Items

Tom: Deliver notification system (2 sprints) — ongoing

Lisa: Pull analytics on onboarding drop-off, share report by Friday

Tom: Set up dev environment for new hire before Monday

Notes

Three enterprise customers reported issues with notification system

New hire joins Monday

Next Meeting

2 weeks

Comparison to Meeting Recording Tools

Tools like Otter.ai, Fireflies, and Fathom record entire meetings and transcribe everything. This approach has real value for detailed record-keeping, but it comes with tradeoffs:

| Aspect | Recording tools | Voice summary with Telvr | |---|---|---| | Effort during meeting | Zero | Zero | | Output type | Full transcript | Structured summary | | Output length | Very long | 1 page | | Processing time after meeting | 10-30 min to review | Under 2 minutes | | Works for all meeting types | Requires joining/recording | Works for any meeting | | Bot in meeting room | Yes (can feel intrusive) | No | | Searchable raw transcript | Yes | No |

The recording approach is better when you need a complete record of what was said — legal contexts, complex technical discussions, or meetings where exact wording matters. The voice summary approach is better for the majority of day-to-day meetings where you need a structured action-item record, not a verbatim transcript.

Practical Workflow Tips

Speak immediately after the meeting. Memory fades fast. The two minutes right after a meeting end are the window when everything is clear. Do not open Slack first.

Walk through the meeting chronologically. Start with "who was there, what was the purpose," then discuss decisions, then actions. A natural narrative structure makes the AI output better.

Name people explicitly. "Tom is going to handle the infrastructure work" produces a named action item. "Someone is going to handle it" does not.

Include deadlines when they exist. "by Friday," "before Monday," and "end of Q2" all get extracted into the action items section.

Use it for informal conversations too. A 15-minute hallway conversation with a manager produces the same kind of action items as a formal meeting. Voice summary works for any verbal exchange that produces commitments.

Integration with Your Workflow

Meeting Notes mode works in any text field, which means the output can go directly into:

Notion or Confluence pages
Email to attendees
Slack channel updates
Linear or Jira comments
Google Docs

The push-to-talk insertion means you do not need to open a special app or export anything. Place your cursor in the destination, speak your summary, and the formatted notes appear.

Why This Works Better Than You Expect

The instinct when people first try voice meeting notes is to worry about missing something. What if you forget a detail between the meeting and the summary?

In practice, the opposite happens. When you are not taking notes during a meeting, you are fully present — you follow the conversation more carefully, you notice the nuances that note-takers miss, and you understand the decisions better. The mental model you build during an attentive meeting is actually more complete than a list of notes taken while only half listening.

The voice summary forces a useful skill: synthesizing what happened into the decisions and actions that actually matter. Full transcripts feel complete but often obscure the key information in a wall of conversation. A two-minute post-meeting voice summary produces more actionable output than a 40-minute transcript.