Sales outreach · Production

Flow AI reduces real estate outreach from days to 60 seconds using n8n and voice AI

The problem

Real estate agents spent three to five hours per day on manual follow-up — cold calling, mass emails, and texting past leads — while needing to comply with strict US regulations including 10DLC and CAN-SPAM. Handling the volume manually required a full-time assistant costing around $50,000 per year.

First attempt

Flow AI's original stack of Zapier, Twilio, and Mailchimp was fragile and unsuitable for a product business: tools were siloed, opt-in status was difficult to enforce consistently, and there was no strong data isolation between customers or centralized logging for audits.

Workflow diagram · grounded in source
1
Agent issues voice command
trigger
“An agent can say, "I am hosting an open house next Saturday, invite everyone in my contacts," and the system takes over”
2
ElevenLabs converts speech
ai_action
“ElevenLabs converts speech into structured input and sends it to n8n via webhook”
3
Database query and compliance check
validation
“workflows query the Postgres database, verify opt-in status and compliance”
4
AI content generation
ai_action
“generate personalized content using AI”
5
Regulatory safeguards applied
validation
“apply safeguards for 10DLC and CAN SPAM”
6
Message routing and delivery
output
“route messages through SMS providers or Mailgun”
7
Audit trail logging
output
“Every step is logged, creating a complete audit trail”
Reported outcome

Campaigns that once took days now run in under a minute via a single voice command.
Agents save hundreds of hours previously spent on manual follow-up, and compliance is built into every workflow. Flow AI went from concept to working product in roughly one month.

Reported metrics
Agent daily follow-up timethree to five hours
End-to-end execution timeunder 60 seconds
Campaign completion time reductioncampaigns that once took days now run in under a minute
Full-time assistant cost avoidedaround 50,000 dollars per year
Show all 7 reported metrics
agent daily follow-up timethree to five hours
end-to-end execution timeunder 60 seconds
campaign completion time reductioncampaigns that once took days now run in under a minute
full-time assistant cost avoidedaround 50,000 dollars per year
time from concept to working productroughly one month
agent time savedhundreds of hours
message delivery time by voice commandwithin about 60 seconds
Reported stack
n8nElevenLabsPostgresMailgunZapierTwilioMailchimp
Source
https://n8n.io/case-studies/flow-ai/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Campaigns that once took days now run in under a minute via a single voice command.

What tools did this team use?

n8n, ElevenLabs, Postgres, Mailgun, Zapier, Twilio, Mailchimp.

What results were reported?

Agent daily follow-up time: three to five hours; End-to-end execution time: under 60 seconds; Campaign completion time reduction: campaigns that once took days now run in under a minute; Full-time assistant cost avoided: around 50,000 dollars per year (source-reported, not independently verified).

What failed first in this deployment?

Flow AI's original stack of Zapier, Twilio, and Mailchimp was fragile and unsuitable for a product business: tools were siloed, opt-in status was difficult to enforce consistently, and there was no strong data isolati…

How is this sales outreach AI workflow structured?

Agent issues voice command → ElevenLabs converts speech → Database query and compliance check → AI content generation → Regulatory safeguards applied → Message routing and delivery → Audit trail logging.

Related sales outreach cases