Browse categories
Explore
Fiverr Pro
English
$
USD
I will build an ai tool that turns voice recordings into structured documents
KyuSung L.
About this gig
You record everything. You organize nothing.
Consultations, meetings, lectures, interviews - piling up in a folder. The problem is not discipline, it is tooling.
I build a pipeline that transcribes your audio (Whisper STT), has an LLM read it, and outputs YOUR EXACT DOCUMENT FORMAT - not a generic summary. Your sections, your terminology, your tone.
PROOF IT WORKS
I built and operate a production SaaS that turns doctor-patient conversations into structured medical chart drafts (SOAP format) - one record button to a chart draft, at paid-product quality. Same pipeline, adapted to your format.
FITS
Consulting (legal, insurance, real estate, therapy), medical and rehab notes, lecture notes, meeting minutes, interview digests. Send me your template and the output matches it.
FREE DEMO BEFORE YOU ORDER
Send one sample recording (5 min or less) and I will return a partial result so you can judge the quality first.
INCLUDED
The tool (or web app) + custom format prompt + docs + a transparent per-document cost breakdown (typically a few cents of AI cost per document).
I work async, in writing: updates and screenshots at every step.
Get to know KyuSung L.
KyuSung L.
Automation and Web Scraping with Python and AI
- FromSouth Korea
- Member sinceAug 2026
Languages
Korean, English
I build automation that replaces someone's daily manual routine - and I'm my own best client: 10+ scraping and alerting pipelines run unattended on my server, for over a year.
Shipped: 14 mobile apps (Flutter) on both app stores, 3 AI SaaS products in production (Whisper + LLM document tools), 10+ data pipelines.
8 years of backend engineering, currently leading a 10-person team. I've put OpenAI and Anthropic APIs into real paid products - I know what survives production.
I work fully async, in writing: daily updates, milestone deliverables, screenshots at every step.
My Portfolio
FAQ
How accurate is the transcription?
Judge it yourself with the free demo before ordering. Domain-specific terms are handled with a custom vocabulary list, which improves accuracy a lot in specialized fields.
Is my audio stored anywhere?
Default design: audio is processed and discarded, never stored. A fully self-hosted setup, where nothing leaves your own server, is available from Standard up.
Can I use this for medical or legal records?
It generates drafts. Final review and sign-off stays with the professional - that boundary is built into the design, not bolted on afterwards.
What languages does it support?
Whisper handles most major languages, including Korean and English. Tell me your language and domain and I will confirm quality with the free demo first.
