Fly Labs
Ask FlyBot

VoiceDraw – Talk system design out loud, the diagrams draw themselves

I was frustated by having to draw system design diagrams by hand when discussing it with my team mates or in an interview. So I thought "Wouldn't it be great if someone draws it for me, while I think out loud?".That is when I came up with VoiceDraw. You can just think out loud or discuss your system architecture with a friend/interviewer, the diagrams are automatically drawn along with your reasoning, open questions and tradeoffs beautifully written on to the side.Demo Video: https://youtu.be/36PgHKSuccE

Hacker NewsToolDesign & CreativeSource
0Sign in to voteCopy link

FL score

62

out of 100

Verdict

VALIDATE

medium confidence

Competition

No competitor data yet

Trend

No signal yet

Speech-to-diagram tool that converts verbal system design discussions into visual architecture documents in real-time.

The pain

Engineers waste time manually drawing boxes and arrows during design discussions and interviews when they should be thinking about the problem. The cognitive load of talking + drawing simultaneously splits attention. Interview candidates especially struggle to articulate architecture while hand-drawing.

The gap

Current tools (Lucidchart, Draw.io, Miro) require manual input. Existing voice tools (Otter.ai, Fireflies) transcribe but don't visualize. No tool bridges the gap between spoken architecture discussion and automatically generated system diagrams with decision tracking.

Build angle

Start narrow: target system design interview prep where the pain is acute and willingness to pay is high. Build a Chrome extension or web app that records a design discussion, transcribes it, extracts components/connections via LLM, and renders a diagram. Validate with 50 interview candidates willing to pay $10-20/month before expanding to team collaboration or enterprise.

Strengths

  • Solves a real, specific pain point in a high-stakes context (interviews)
  • Demo shows working prototype, not vaporware
  • Low barrier to entry for users (just talk naturally)
  • Defensible moat if accuracy/speed becomes significantly better than manual tools
  • Clear monetization path (B2C subscription, B2B licensing to interview platforms)

Risks

  • Speech-to-diagram accuracy heavily dependent on LLM quality; messy real conversations may produce garbage diagrams requiring manual cleanup anyway
  • Market size unclear, how many people actually struggle with this enough to pay? Interview prep is seasonal and one-time purchase behavior
  • Existing tools (Lucidchart + voice notes, or just Excalidraw) may be 'good enough' for most users
  • Requires continuous LLM API costs; unit economics may not work at low price points
  • Competitive threat from Figma, Miro, or Claude adding voice features natively
  • Diagram quality and layout are hard problems; auto-generated diagrams often look worse than hand-drawn ones

Questions about this idea?

FlyBot reads the scoring and gives you a second opinion on “VoiceDraw – Talk system design out loud, the diagrams draw themselves”.

Open FlyBot