Speaker Memory — Every Person You Have Ever Spoken To, Remembered
What it is
Speaker Memory is auraScribe's "Remember" layer — a persistent database of every person you have ever spoken to in a recorded meeting. Each speaker accumulates a profile over time: voice samples, signature quotes, role and company history, DISC interaction profile averaged across sessions, and the full cross-meeting context of every conversation you have had with them. The next time the same person appears in a meeting, the system recognises them and brings the entire history with them.
Why it matters
Every meaningful professional conversation is shaped by the conversations that came before it. You should walk into the second meeting with someone armed with everything you learned in the first. In practice, that history lives in your head, gets fuzzier each week, and is gone within months. Speaker Memory makes the history permanent and searchable. When you meet someone again — six weeks later, six months later — the system surfaces what you discussed, what they care about, how they communicate, and what they committed to, so you can pick up where you left off instead of starting over.
How auraScribe does it
After Pass 2 profiles speakers in a meeting and you confirm or correct their identities during the review stage, the speaker database is updated. New speakers get a fresh profile with their voice sample, signature quotes, role, and company. Returning speakers have their profile enriched — new voice samples are added, DISC interaction profiles are averaged across sessions, and the new meeting is linked into their history. The next time their voice appears in a recording, the system recognises them via your confirmed speaker data and pulls their full profile into the analysis, including all prior commitments tracked under their name.
Who it's for
- Sales professionals managing relationships with prospects, champions, and customers across many meetings
- Consultants serving recurring clients whose context spans months of engagements
- Coaches tracking client communication evolution across an entire coaching arc
- Managers building deeper understanding of each team member over time
Frequently Asked Questions
Is this voice biometric recognition?
No. Speaker Memory does not use voice biometrics to identify speakers in new recordings. It uses the standard 2-pass identification (acoustic diarisation plus transcript-based name deduction) and is enriched by your confirmations from the review stage. The database grows from your verified data, not from biometric matching.
What is stored about each speaker?
Voice samples extracted from prior meetings, signature quotes that capture how they speak, their role and company history as identified across sessions, a DISC interaction profile averaged from their behavioural patterns across meetings, and links to every meeting they have appeared in. All of this is yours alone — no one else can access your speaker database.
Can I edit or delete speaker profiles?
Yes. You control your speaker database entirely. You can edit names, roles, companies, merge duplicate profiles, delete individual speakers, or wipe the database. The database is your private record of the people you have spoken to, not a public profile of them.
How is the DISC profile produced?
The DISC interaction profile is averaged from the behavioural observations captured in Raw Audio Cues across every meeting the speaker has appeared in. It reflects how they consistently communicate in your conversations — not a personality assessment they have taken. As you accumulate more meetings with the same person, the profile becomes more stable and more useful for tailoring your next interaction.