Alma Eriksson
View profile
Inbound-meeting router
Calendly bookings classified by intent via GPT-4, routed to the right AE in Slack, full context attached.
No-show rate down 22%; first-call quality up.
0
Problem
Inbound Calendly bookings landed on whoever was on duty — half the time the wrong AE, no context, awkward first 10 minutes. Prospects dropped off.
Flow
Calendly webhook → GPT-4 classification → round-robin within target pod → HubSpot update → Slack DM 15 min before call.
Cadence
Per inbound booking, ~12/day
What I built
- Calendly webhook trigger on each new booking
- GPT-4 classifies intent from the booking notes + lead source
- Round-robin within the right AE pod (mid-market vs SMB vs enterprise)
- Slack DM to assigned AE with full lead context 15 min before call
Limits
- English-only intent classification — non-EN bookings go to default rep
- Doesn't reschedule — only routes; reschedule is on Calendly itself
AI tools used
- OpenAI API
- Zapier
- Slack
@alma.eriksson
