An AI intake assistant that filters 60% of unqualified enquiries
A boutique firm was spending paralegal hours on consultations that were never going to become cases. We put qualification in front of the calendar.
- Scenario
- Example: boutique law practice
- Industry
- Professional services · Chicago, IL area
- Timeline
- 4 weeks including compliance review
60%
of unqualified enquiries filtered pre-call
22 hrs
paralegal hours returned monthly
+47%
consult-to-retainer rate
24/7
coverage, no answering service
The problem
Roughly half of booked consultations fell outside the firm's practice areas or jurisdiction.
Intake notes were inconsistent, so attorneys walked into calls unprepared.
Enquiries arriving overnight waited until business hours for any acknowledgement.
What we built
01
Scoped assistant
A site assistant trained strictly on the firm's practice areas, jurisdictions, and fee structure — with an explicit hand-off when it isn't sure.
02
Structured intake
Every conversation produces a standardized case summary in the CRM before an attorney sees it.
03
Human where it matters
Qualified matters route straight to a real calendar slot; edge cases escalate to a person rather than guessing.
Stack: Lovable AI Gateway · Clio · Calendly · Slack
What changed
- Attorneys now spend consultation time on matters that can actually convert.
- Overnight enquiries get a substantive response immediately instead of a voicemail box.
- Intake quality is consistent enough to report on for the first time.
“It says 'I'm not certain, let me connect you' instead of inventing an answer. That's the only reason we shipped it.”