// transcription for journalists
You have a 45-minute interview recording and a 2-hour deadline.
The source said something important at the 30-minute mark but you cannot remember the exact phrasing. You need the direct quote, not a paraphrase. You need it attributed to the right speaker. And you need to file the story before lunch. This is the problem SHRP solves. Upload the recording, get a transcript with speaker labels and timestamps, find the quote, and get back to writing.
// the workflow
There are three steps and no learning curve.
Upload your file. Drag your MP3, M4A, WAV, MP4, or any common audio/video format into SHRP. Files up to 500MB on Pro. Most phone recordings and digital recorders work without conversion.
Wait a few minutes. AssemblyAI processes your audio on their servers. A 45-minute interview typically takes 3-5 minutes. You can close the tab and come back later — the transcript will be waiting.
Read, search, export. Your transcript appears with speaker labels and timestamps. Use Ctrl+F to find specific words or phrases. Copy to clipboard, download as .txt, or export as SRT/VTT if you need subtitles.
// what you get
The transcript is not a wall of text. It is structured so you can actually work with it.
Speaker labels. Each change of speaker is marked — Speaker A, Speaker B, and so on. In a two-person interview this means you can instantly distinguish your questions from the source's answers. In a panel or press conference with multiple voices, SHRP detects and labels each speaker automatically.
Timestamps. Every utterance is timestamped so you can jump back to the audio to verify a specific quote. When your editor asks "did they really say that?" you can find the exact moment in seconds.
Confidence highlighting. Every word is color-coded by how confident the AI is in its transcription. High-confidence words appear normally. Lower-confidence words are highlighted in yellow or red. This tells you exactly where to double-check against the audio — instead of re-listening to the entire recording.
AI extraction. One click to pull out a summary, key names and dates, or action items. Useful for long-form interviews where you need to identify the main threads quickly.
// accuracy
SHRP uses AssemblyAI, which reports around 98% accuracy on clean audio in English. In practice, your results depend on recording quality. A quiet room with a decent microphone will give you near-perfect transcripts. A phone recording in a noisy cafe will not.
Where it works well: one-on-one interviews in a quiet setting, press conferences with a good PA system, phone calls recorded through an app, Zoom recordings with clear audio.
Where it struggles: heavy accents the model has not seen often, overlapping speech where multiple people talk simultaneously, highly technical or domain-specific jargon (legal terms, medical terminology, niche policy language), and poor-quality recordings with heavy background noise. In these cases, treat the transcript as a first draft and verify critical quotes against the audio. The confidence highlighting helps you know where to look.
// export and file
Copy the full transcript to your clipboard in one click. Paste it into Google Docs, Word, your CMS, or wherever you write. You can also download as a .txt file, which works with any text editor.
If you produce video or audio packages, export as SRT or VTT for subtitles. These formats include timestamps and work directly with video editing software and web players.
For newsrooms that need to archive transcripts, save to My Projects inside SHRP. Projects sync across devices if you have an account. Search your past transcripts by keyword when you need to reference an old interview months later.
// pricing for newsrooms
Voice typing with a microphone is free, forever. No limits, no account required. Useful for dictating notes or drafting stories by voice.
If you only transcribe a handful of interviews per month, pay-as-you-go credits are the most economical option. If you transcribe regularly, Pro pays for itself within a few uses.
Upload your next interview recording and see the transcript in minutes. No account required to try voice typing. File upload needs a Starter or Pro plan.