Speaker-labelled transcript
Timestamped text with anonymous speaker turns, which you can name during review. Voice identity matching is not part of the first version.
AI meeting suite
From conversation to approved next steps.
The intelligence layer turns a supported recording into a structured meeting record: transcript, summary, decisions, tasks, deadlines, open questions and customer requirements — then waits for a person to approve it.
The workflow
Seven steps, and a person at the end of every one.
This is the process the software is built around. It runs after a meeting, on audio the team is allowed to record.
Step 1 · Consent
Participants are informed and consent is recorded before capture starts. A visible indicator shows the meeting is being recorded.
Step 2 · Capture or import
The room speaker captures the meeting, or the team uploads a recording. The original audio is preserved untouched.
Step 3 · Transcribe and label
Speech is converted into a timestamped transcript, speaker turns are labelled anonymously, and language hints are added per segment.
Step 4 · Extract
Topics, decisions, commitments, tasks, owners, deadlines, open questions and customer requirements are pulled out in structured stages.
Step 5 · Check against the transcript
Every extracted item is checked against the words that support it. Ambiguous or unsupported items are flagged or withheld.
Step 6 · Review and correct
A person replays the relevant audio, fixes errors, names the speakers and approves the record.
Step 7 · Export or approved send
Approved outputs are exported, or a follow-up message is sent after review — never automatically.
What you get
Everything a meeting is supposed to leave behind.
Each output is designed to be reviewed and corrected. None of it is generated as a final answer.
Summary and topics
A short account of what was discussed, built from validated items rather than invented detail.
Decisions and commitments
What was agreed, with a link to the transcript passage that supports it.
Tasks with owners and dates
Action items with an owner and due date when they were clearly stated; uncertain values are flagged instead of guessed.
Open questions
Questions that were raised and not answered — the list most meetings never write down.
Customer requirements
Requirements, budgets, delivery preferences and commitments from customer meetings, with status and evidence.
Review workspace
Replay audio, correct text, rename speakers and approve — with the correction history kept.
Exports
Approved outputs as Word, PDF or plain text.
Follow-up drafts
A draft for email or WhatsApp in the right language and tone, ready for a person to review and send.
Evidence
Evidence for every claim.
Click a decision to see the transcript lines behind it. If the AI is unsure, it flags the item instead of guessing.
Illustrative example. Real outputs are generated from the customer’s own recordings during review.
Speaker 1We can go live with the Pune warehouse in the first week of next month.
Speaker 2That works, but the bulk pricing needs sign-off from our CFO before we commit.
Speaker 1Understood. I’ll send the revised numbers tomorrow morning.
Speaker 3Should we include the Gujarat depot in the same rollout?
Companion app
Review and approve from anywhere.
The mobile app extends the same workspace beyond the meeting room. It is for meetings you start yourself, and for reviewing what was already captured.
- Start, pause and stop a recording you initiated, with a visible recording state.
- Import a recording or transcript from a supported meeting tool.
- Read the summary, correct speakers and approve outputs.
- Get reminders for approved tasks and due dates.
Always user-controlled
Capture requires consent and device permissions. There is no always-on recording, and no promise of automatic call capture — mobile operating systems restrict it, and we will not claim otherwise.
What is not in the first version
- Live transcription during the meeting
- Voice identity recognition
- Cross-meeting semantic search
- Broad CRM and calendar integrations
- Translation and spoken summaries
These are roadmap items. See the roadmap