Everything it takes for a customer interview to still count in three years
Capture, understand, ask. Every feature is included in every plan.
Every conversation gets in, wherever it comes from
On-site recording
- Record in the browser on phone, tablet or laptop, as an installable web app
- Recordings are saved in chunks and keep going if the connection drops
- Capture metadata such as date, customer, segment and interviewer right away
Upload and bulk import
- Upload audio, video and transcripts (VTT, SRT, DOCX, TXT)
- Many files at once, such as recordings from Google Meet or an entire legacy archive
- Status per file, with a clear reason if a file fails
Import from Microsoft Teams
- Pick and import your own Teams meetings from the last 30 days that have a recording or transcript, no bot in the meeting
- The recording is re-transcribed via AssemblyAI in the EU; speaker names from Teams show up as suggestions
- One-time approval by a workspace admin; your Microsoft 365 admin enables transcript access
Who said what, spelled correctly
German and English transcription
- Transcription via an EU-based service (AssemblyAI), including speaker separation
- Optionally set the number of speakers for more accurate separation
- Transcript scrolls in sync with the player; click to jump to any moment
Speaker separation and recognition
- Speakers are separated automatically and labeled as interviewer or customer
- Name a person once and Discoverify recognizes them in other recordings
- Matches for team members appear as suggestions; every assignment can be undone
Product glossary
- Maintain product areas, screens and modules with synonyms and common mishearings
- Improves the spelling of technical terms in the transcript
- Maps statements to product areas automatically, no mandatory tagging
Answers you can trust
Ask across all interviews
- Ask in your own words, including follow-up questions in the same thread
- Points with number of customers, time range, segments and trend, plus individual mentions
- Active filters are always visible: time range, customer, segment, person, product area
Verbatim quotes with audio
- Every quote appears word for word in the transcript
- One click plays the moment right inside the answer
- Without evidence, the answer is “not found”, not a guess
Search across time
- Ask about time ranges: “What’s new since 2025?”
- See per point whether it’s growing, stable or mentioned less often
- Topic pages with statements per quarter
Key takeaways and topics
- A summary per interview with pain points, requests and praise, each with evidence
- Topic pages per product area with top quotes and a list of customers
- Copy answers with sources or share them as a link
Your data, your tools
Claude via MCP
- MCP server for Claude Code and Claude Desktop
- Personal access tokens, read-only and revocable at any time
- Same answers and evidence as in the web app
Export and deletion
- Full export as a ZIP: metadata as JSON plus media files
- Delete interviews along with all derived data
- Admin and member roles
Passwordless sign-in
- Sign in with a magic link sent to your email
- Or with your Microsoft or Google account, including at sign-up
- No password to reuse or guess
Data stays in Germany and the EU
- Hosted by Hetzner in Germany
- Claude and embeddings via Amazon Bedrock in the EU
- Transcription via AssemblyAI in the EU; voice profiles stay on our server and are never exported
What we’re working on
These features don’t exist yet. We won’t name a date until they’re ready.
- Answers in the language of your question (today: German)
- Import from Google Meet
- Automatic Teams import for meetings flagged in advance
- Connector for claude.ai