Confidential · © 2026 Space-O Technologies Inc. · Prepared exclusively for DMA. Please do not copy, save or share.

Section 20 · Sales Rep Work Elimination / Zero-Entry CRM

20.1Automatic call/meeting summary

Built in the POC
No.
20.1
DMA's reference
Decision Book p.35
Phase
Phase 1
Autonomy
L3 · Act within rules
DMA's decision
Yes

★ Marked "non-negotiable, build first" in the book. ◆ Amended in v3 of the book.

What DMA asked for · Decision Book p.35

After every call or meeting, write the summary into the CRM so the rep doesn't have to.

Why it matters to DMA: Reps skip notes when they're busy; the CRM then forgets what the prospect said. This is the book's first build item for section 20.

In our words, from DMA's Product Decision Book v3, page 35.

How we do it

  • The rep opens the account and pastes the transcript, or drops the Zoom transcript file. That is the only step.
  • The fast model reads the call once and returns a summary, the facts with exact quotes, the promises on both sides and the follow-up points. Code checks every quote against the transcript and drops anything it can't find.
  • The summary becomes a Nutshell note, waiting with the rest of the call's changes as one batch for one approval.

The flow

It plays on its own while it's on screen; hover or use the controls to pause or step through.

Transcriptpasted or a Zoom fileMatchcompany and dealReadsummary, facts, promisesChecksquotes, dates, amountsOne batchwaiting for approvalNutshellnote on the deal
Step 1 of 6

The call

Tom pastes the Summit HVAC call on the account.

  • Person
  • Rule in code
  • AI
  • Screen
  • Nutshell

Where the data goes

The same six stages on every page. Nothing reaches Nutshell except through the write gate.

  1. 1Call transcript

    Source

    A call transcript: pasted, or a Zoom .vtt / .txt file dropped on the account (text only, 512 KB at most).

  2. 2Sync

    Copies Nutshell changes into the bridge database

    Reads the account, people and open deals from the mirror, to match the call to the right company.

  3. 3Bridge database

    A copy of the CRM data, plus what the AI works out

    The run, the facts with their quotes and the e-mail draft are stored; the transcript is kept encrypted.

  4. 4AI

    Claude models, only through DMA's own gateway

    The fast model reads the call and copies exact words; code checks every quote, date and amount. The conversation model drafts the follow-up.

  5. 5Write gate

    The only way back into Nutshell: checked, approved when needed, sent once

    Every change waits as one batch for one approval, then leaves through the write gate with 30 seconds to undo.

  6. 6Nutshell

    The system of record

    The note, the tasks and the fields appear on the deal, marked as written by DMA AI.

The tables behind it

Drawn from the POC's database catalogue: structure only, no data.

ai.scribe_runsPKiduuidtitletextmeeting_datedateFKconversation_iduuidstatustextaccount_iduuidFKapproval_iduuidfollow_up_draft_iduuid+ 16 more columnsctrl.approvalsPKiduuidkindtextstatetextsummarytextdecided_bytextundo_untiltimestamp+ 16 more columnsctrl.outboxPKiduuidoperationtextbefore_imagejsonbautonomy_leveltextFKapproval_iduuidapproved_bytextstatetext+ 21 more columnsai.extracted_factsPKiduuidfact_keytextvaluejsonbconfidencenumericevidence_quotetextobserved_attimestampstatustextIDprompt_iduuid+ 7 more columnsconv.email_draftsPKiduuidIDconversation_iduuidkindtextsubjecttextchecksjsonbstatustextIDapproval_iduuidIDagent_version_iduuidIDdecision_iduuid+ 30 more columns

PK primary keyFK reference the database enforces (solid line)ID reference kept by id (dashed line)

TableWhat it holdsColumns
ai.scribe_runsOne row per logged call: the transcript's fingerprint, the account it matched, the extraction and the approval batch it produced.24
ctrl.approvalsThe approval queue: one row per batch waiting for a person, with its risk and the undo window.22
ctrl.outboxThe write gate: each proposed Nutshell change, its approval and what it replaced.28
ai.extracted_factsFacts the AI found, each with the exact quote it came from and a confidence.15
conv.email_draftsE-mail drafts, encrypted, with the checks they passed; a draft is never sent without approval.39
How the tables connect (8)
ColumnPoints toKept by
ai.scribe_runs.conversation_idconv.conversationsThe database (foreign key)
ai.scribe_runs.approval_idctrl.approvalsThe database (foreign key)
ctrl.outbox.approval_idctrl.approvalsThe database (foreign key)
ai.extracted_facts.prompt_idagent.promptsThe application (by id)
conv.email_drafts.conversation_idconv.conversationsThe application (by id)
conv.email_drafts.approval_idctrl.approvalsThe application (by id)
conv.email_drafts.agent_version_idagent.agent_versionsThe application (by id)
conv.email_drafts.decision_idctrl.decision_logThe application (by id)

Worked example

Synthetic demo data: every name, business and number is made up.

A logged call: the screen and the system

1 / 3

What you see

1.Every change, with its evidence

The call summary note and the tasks, each with the words it came from.

localhost:13301/approvals
The CRM updates from one logged call, waiting for one-click approval (a real run on synthetic data).
Callout 1

What happens in the system

  • Conversation servicescribe

    The transcript is checked as text only, matched to Summit HVAC by its attendees, and read once.

  • LLM gatewayclaude-haiku-4-5

    3,249 tokens in, 653 out, $0.0065: it copies exact quotes; it never writes a value of its own.

  • Rule enginequote, date and amount checks

    Code confirms each quote is really in the call, and turns "Tuesday" into 29 September from the meeting date.

A real run of the scribe on a synthetic call.

One logged call: "Summit HVAC call" (synthetic call, real run)

Matched to
Summit HVAC · deal "Summit HVAC - Local + PPC"By the attendees and the company name; no model involved.
Read by
claude-haiku-4-5: 3,249 tokens in, 653 out, $0.0065
Follow-up drafted by
The conversation model, then the claims check: $0.0056
Time
About 12 seconds from paste to the approval batch
Waiting for approval
1 note, 2 tasks, 5 facts kept on the account, 1 follow-up e-mail draft

In the running POC

localhost:13301/approvals
The CRM updates from one logged call, waiting for one-click approval (a real run on synthetic data).

The CRM updates from one logged call, waiting for one-click approval (a real run on synthetic data).

  1. 1Each change and the exact words it came from.
  2. 2The follow-up e-mail, drafted and claim-checked; approving saves it, it never sends.
  3. 3One click approves the whole batch; each line can also be edited or dropped.

Dashboard · Approvals, the CRM updates from one call transcript. Captured from the running POC on September 25, 2026; synthetic data.

See it live

Opens the exact screen in the running POC, in a new tab.

OpenWhereWhat to try
Accounts → Log a callDMA AI dashboardOpen an account and paste or drop a Zoom transcript.
Approvals → CRM updatesDMA AI dashboardThe changes from a logged call, each with the quote it came from.

Status

Built in the POC

Built in the POC. This capability runs in the POC today, on synthetic data, end to end.

20.1 Automatic call/meeting summary · DMA AI proof