Just built a Slack bot that posts key metrics from daily standup transcripts. Code snippet inside.

Hey everyone! I’m so excited to share a little weekend project I just wrapped up that’s already making my marketing ops life easier. As many of you know, I’m a huge fan of Fireflies.ai for capturing meeting transcripts—it’s been a game-changer for our daily standups with the sales and marketing teams. But I found myself wanting a quicker, more visible way to track key outcomes and action items without having to dig through the full transcript every time.

So, I built a simple Slack bot that automatically parses our daily standup transcripts from Fireflies and posts a formatted summary into our designated #standup-metrics channel. It pulls out things like:
- **Key decisions made** (e.g., "Pause campaign X," "Move forward with integration Y")
- **Blockers flagged** by team members
- **Action items** with owners (using simple pattern matching for "I will," "assigned to," etc.)
- **Sentiment tone** (just basic positive/neutral/negative from keywords—nothing too fancy!)

The bot runs on a schedule every morning, fetches the transcript from Fireflies via their API (for the previous day's standup), processes the text, and then uses Slack's webhook to post. It’s built in Python and hosted on a simple AWS Lambda. Here’s the core snippet for the parsing logic—it’s pretty straightforward but effective for our structured standup format:

```python
# (Example snippet - I'd share the full repo link if anyone's interested!)
def extract_highlights(transcript):
highlights = {
'decisions': [],
'blockers': [],
'actions': [],
'tone': 'neutral'
}
# Simple line-by-line checks for keywords
for line in transcript.split('\n'):
line_lower = line.lower()
if 'decided' in line_lower or 'agree' in line_lower:
highlights['decisions'].append(line.strip())
if 'block' in line_lower or 'stuck' in line_lower:
highlights['blockers'].append(line.strip())
if 'i will' in line_lower or 'assigned to' in line_lower:
highlights['actions'].append(line.strip())
# Basic tone check (very simple version)
positive_words = ['great', 'progress', 'yes', 'done']
negative_words = ['concern', 'issue', 'waiting', 'can’t']
# ... scoring logic here
return highlights
```

This has been awesome for keeping the entire revenue team (marketing + sales ops) aligned without extra manual work. I’m curious—has anyone else built custom integrations or automations on top of Fireflies data? I’d love to swap notes on how you’re piping meeting insights into your CRMs (we use HubSpot) or analytics dashboards. Maybe there are better ways to handle the natural language processing part for action items!