COS Pipeline

After Install

Setup is done. Here's where your documents live and how to connect your recording service.

1
Open your dashboard
The URL was printed at the end of the installer

Your dashboard is running locally. Open it in your browser:

Log in with the username and password you set during the installer. If your port is different, the installer printed it — look for the line that says Dashboard : http://localhost:XXXX.

On iPhone or another device: use the Tailscale URL printed at the end of Step 3 during install (http://100.x.x.x:7777).
2
Find your Drive folder IDs
You'll need these to connect your recording service

Run this command in Terminal. It reads your config and prints every folder ID the pipeline is watching:

Terminal
python3 -c " import yaml, os cfg = os.environ.get('COS_CONFIG_DIR', os.path.expanduser('~/cos-pipeline-config')) ctx = yaml.safe_load(open(f'{cfg}/firm_context.yaml')) for src in ctx.get('transcript_sources', []): print(f\"\n{src['name']}\") for fid in src.get('folder_ids', []): print(f\" https://drive.google.com/drive/folders/{fid}\") "

This prints one clickable Drive URL per folder. Open each one to confirm it exists in your Google Drive, then copy the folder ID (the long string at the end of the URL) to paste into your recording service in the next step.

Can't find the folder? The installer creates these folders automatically. If one is missing, re-run ./setup.sh --instance=<your-slug> --validate to diagnose.
3
Your Google Docs — where everything gets written
Created automatically during install — start empty, fill up over time

Run this to print direct links to all three of your shared docs:

Terminal
python3 -c " import yaml, os cfg = os.environ.get('COS_CONFIG_DIR', os.path.expanduser('~/cos-pipeline-config')) docs = yaml.safe_load(open(f'{cfg}/config/drive-docs.yaml')) for k, v in docs.get('docs', {}).items(): print(f\"{k}: https://docs.google.com/document/d/{v}\") "
Follow-ups
Action items from every email and call, tagged by owner
Written by pipeline
Pipeline / Deal Intel
Full investor memos from every processed call
Written by pipeline
People / CRM
Contacts extracted from calls and emails
Written by pipeline
Bookmark all three. The pipeline writes to them automatically — you read, it writes.
4
Connect your recording service to Drive
One-time setup — after this, every call is processed automatically

Paste the folder ID from Step 2 into whichever service you use. Only need to do this once.

Otter AI
Requires Zapier
  1. Go to zapier.com → Create a new Zap
  2. Trigger: Otter AI — New Recording
  3. Action: Google Drive — Upload File
  4. Set the destination folder to your Otter AI folder ID from Step 2
  5. Turn the Zap on — every new Otter recording deposits automatically
Beside AI
Native Drive sync
  1. In Beside AI, open Settings → Integrations → Google Drive
  2. Connect your Google account
  3. Set the export folder to the folder ID from Step 2
  4. Enable Auto-sync after each meeting
Fireflies
Native Drive sync
  1. In Fireflies, go to Integrations → Google Drive
  2. Connect your Google account
  3. Select the transcript folder from Step 2 as the export destination
Grain
Native Drive sync
  1. In Grain, go to Workspace Settings → Integrations → Google Drive
  2. Enable Drive export and select your folder from Step 2
Read AI
Native Drive sync
  1. In Read AI, go to Settings → Integrations → Google Drive
  2. Enable sync and point at your folder ID from Step 2
Fathom
Native Drive sync
  1. In Fathom, go to Settings → Integrations
  2. Enable Google Drive sync and select your folder from Step 2
Manual transcript
No setup needed
  1. Paste your transcript text into a new file and save it as Call Title.txt
  2. Upload it to any of your configured Drive folders from Step 2
  3. The pipeline picks it up at 7:22am or immediately via the real-time webhook
5
What runs and when
All automatic — nothing to trigger manually
6:45am — Market intelligence fetch 7:22am — Transcript backfill 7:51am — Personal briefing Every 2h — Email triage Always on — Dashboard server Real-time — New transcript in Drive

To trigger any pipeline manually or check when it last ran, go to your dashboard admin panel: