A consultant's day
8 to 10 client meetings a day, a recap to write up after each, 3 projects in parallel to push between meetings. You brief agents for research, memo drafts, deck outlines. The keyboard at 62 wpm becomes the bottleneck.
// USE CASE — CONSULTING
Transcribe meetings. Brief agents. By voice.
Dictate in any app on macOS and Windows; on macOS, also capture every client meeting as structured text to ship deliverables 2× faster.
video placeholder · 16:10
// IN SHORT
Dikton is voice dictation for consultants on macOS and Windows, with meeting transcription on macOS. On Mac, during a meeting, Dikton transcribes system audio live, separates speakers, and gives you text ready to paste into Claude or ChatGPT for the recap. Between meetings, hit Alt+D (Option+D on Mac) and brief your AI agents in seconds — a 200-word brief takes a minute. Custom profiles adapt tone and structure to the target app: research, memo, deck, follow-up.
Capture the meeting
On macOS, Dikton transcribes system audio live (Zoom, Meet, Teams, Webex). Speakers separated. At the end of the meeting, copy the text, paste it into Claude — the recap, decisions, and action items come out in 30 seconds instead of 30 minutes.
Brief agents between meetings
Stack 3 briefs in 5 minutes: competitive research in Claude, memo draft in ChatGPT, deck outline in Notion. For each target, a profile with your custom prompt (format, translate, tighten). Switch apps, Dikton applies the right profile before pasting at the cursor.
Ship deliverables
Gmail follow-ups, Notion memos, Figma deck comments — dictated by voice, profiled per app. The personal dictionary knows your clients, markets, frameworks, team names. The recap that took 30 minutes is now done in 5.
// QUESTIONS
Before you ask.
How does Dikton transcribe a client meeting?
On macOS, during the meeting (Zoom, Meet, Teams, Webex), Dikton records system audio and transcribes live. Speakers are separated. At the end, copy the structured text and paste it into Claude or ChatGPT for the recap and action items.
Does the client need to install or allow anything?
No. Dikton captures system audio on your machine — what you hear is what gets transcribed. No client-side action, no permission to request.
What about transcript privacy?
Processing goes through a secure cloud. Audio is processed, never stored — zero retention. Only anonymous analytics (no content) are kept.
How do I brief 3 agents in parallel?
Create a profile per agent or per use: research, draft, deck. Each profile targets an app (Claude, ChatGPT, Notion). Dictate in the target app, Dikton applies the right profile before pasting at the cursor.
Does it work for all virtual meeting types?
Yes, on macOS — anything that goes through system audio: Zoom, Google Meet, Teams, Webex, FaceTime, Slack Huddles. No platform-side configuration.
Put down the keyboard. You don't need it.
Free and open source. Bring your own keys — no card, no subscription.