Automation for advisors · Runs after a meeting is transcribed
Client next steps
A running log of client action items and due dates from call transcripts. Your assistant runs it after a meeting is transcribed and tells you when something needs your attention.
Client next steps — Sep 17
Automation run • Sep 17
Client next steps
Output
I processed the call transcript from your cash forecast review with Harbor & Pine. Four action items have been added to your running next-steps log in Google Workspace.
Harbor & Pine: Cash forecast review
- Date: September 17, 2026
- Attendees: Rhea Quill, Maya Calder, Theo Marsh
| Owner | Action item | Due date | Status |
|---|---|---|---|
| Rhea Quill | Revise September cash forecast to include updated inventory payments to Mossbank Textiles | Sep 21, 2026 | Logged |
| Theo Marsh | Send the August wholesale collections report | Sep 18, 2026 | Logged |
| Maya Calder | Review proposed payment schedule for Fernwick Packing | Sep 23, 2026 (estimated) | Logged |
| Rhea Quill | Prepare variance summary ahead of October board review | Sep 28, 2026 (estimated) | Logged |
Due dates for Maya Calder and your variance summary were not explicitly stated during the call, so I marked them as estimates based on next week's review cycle. Please let me know if you want to change either date.
Your running log now holds four open items for Harbor & Pine. Nothing was sent or shared with client attendees.
Automation steps. Your assistant follows these 3 steps each time it runs, using your connected email and calendar.
- 01
After each client call, read the transcript and pull out every action item and next step with its owner and due date, inferring a date where none was said and marking it an estimate.
- 02
Attribute each to the right client and engagement, and ask rather than guess when unsure.
- 03
Add them to my running next-steps log so there's one place that always shows what's outstanding.
Adjust the automation in chat
Tell your assistant what to do differently and it sets up that version for you.
- Only log action items assigned to me or Quill Advisory.
- Flag any inferred due date in bold text so I can review it quickly.
- Separate next steps by client engagement instead of keeping one combined table.
Results arrive as a message from your assistant. Each run posts to your conversation with your assistant, and can email you too. The automation keeps a log in your knowledge base and reads it before each run to see what has changed.
What starts this automation
Runs after a meeting is transcribed.
Only useful reports
When nothing needs attention, the run says so in a line or stays quiet. When there are findings, you get them with a link to the full run.
What it does on its own
Pulls action items and updates your running log on its own. Sends nothing. An automated run has no tool that sends, posts, shares, or deletes anything.
This automation reads your email and calendar. Connect Google Workspace or Microsoft 365 to Fieldwork. The automation uses whichever you connect. It also uses your files, meeting notes when connected.
More automations for advisors.
Questions about this automation.
- How do I set this up?
- Go to the Agents page in Fieldwork, open your assistant, locate this automation in the job library, and select Add. If your calendar or transcription tools like Granola or Fireflies are not connected yet, the setup card prompts you to link them first.
- What does it do on its own versus what waits for me?
- It pulls action items from each transcript and adds them to your running next-steps log in your Fieldwork knowledge base. It never messages participants, shares documents, or assigns tasks in other apps.
- Where will the results arrive?
- A summary of newly added action items appears in your conversation with your assistant right after the transcript finishes, and it can be emailed to you as well. If a call has no action items, the assistant notes that briefly or stays quiet.
- What happens if a due date is not mentioned on the call?
- The assistant infers a reasonable date from the conversation and labels it as an estimate in your log. To change one, edit the log or ask your assistant in chat.
Security first
No training on your data
Neither we nor our model providers train on your data. Inference runs with zero retention wherever the provider supports it.
Access
Agents only see what the person asking can see. Conversations, files, apps and connections stay private until shared, and admins have no way into conversations.
Encryption and isolation
Data is encrypted at rest and in transit. Each user's code runs in its own microVM, and secrets reach models by name only.
Continuity
Continuous backups, replicated to a second region. Data is exported on request and deleted within 30 days of a contract's end.