ClarityScope Optometry — Appointment & Insurance Automation
A 4-location optometry chain turned six hours a day of front-desk phone work into a tuned AI agent — without replacing their existing PMS.
Impact
What changed.
40% fewer front-desk calls
Call volume across the 4 branches dropped 40% within the first 30 days. Front desk staff were redeployed to in-person patient experience and frames-floor support — zero layoffs, measurably better patient feedback.
Pre-auth in under 24 hours
Insurance pre-auth turnaround dropped from a 3-day average to under 24 hours. Exceptions hit a shared queue instead of disappearing into a single staffer's inbox.
Inside their PMS, not around it
The agent runs inside their existing practice management system. No rip-and-replace, no new SaaS subscription, no separate dashboard for the team to learn. Documentation handed over to their internal IT.
The challenge
Before
ClarityScope Optometry was running four branches under one brand, and every front desk was buried in the same three jobs: confirming appointments by phone, chasing insurance pre-authorisation documents, and following up on reminder no-shows. Each branch was losing roughly six hours a day to phone-and-paper work — call-backs, voicemail tag, scanning insurance forms, and emailing carriers for clarification. Leadership didn't want to rip out their existing practice management system or hire four more receptionists. They wanted one focused automation that played nicely with what they already had.
- ~6 hours/day per branch lost to appointment confirmation calls
- Insurance pre-auth documents printed, scanned, and emailed manually
- No-show rate dragging down per-branch revenue
- Reminder follow-ups handled inconsistently across the 4 locations
- Front desk staff working through lunch to keep up with calls
- Insurance pre-auth turnaround averaging 3 days
- No central queue for exceptions — each branch handled its own escalations
- Existing PMS in place — no appetite for a rip-and-replace
The solution
What we built
We ran our 4-week Business Automation Package. Week one we co-decided the scope: appointment confirmation and insurance pre-auth — the two workflows draining the most hours, run as one connected agent. We built directly into their existing PMS rather than around it. The agent: (1) confirms appointments via SMS and email, with two-way reply handling, (2) classifies insurance pre-auth documents arriving by email and scan, extracting the carrier, plan, and pre-auth status, and (3) escalates anything outside its confidence threshold to a shared human queue — no silent failures. Friday demos every week. By week four it was running against real patient traffic across all four branches. The 30-day post-launch tuning surfaced about a dozen edge cases on insurance plans, all resolved before handover.
Core workflow connections
How the system flows.
- Appointment booked in PMSAI agent sends SMS + email confirmation
- Patient replyAI parses intent (confirm / reschedule / cancel)updates PMS
- Insurance pre-auth document in inboxAI classifies carrier, plan, status
- High-confidence pre-authauto-filed against patient record in PMS
- Low-confidence documentrouted to shared human exception queue
- Reminder cadenceAI agent triggers 7-day, 24-hour, and 2-hour touches
- Branch dashboardlive view of confirmations, exceptions, and pre-auth status
- 30-day post-launch tuning12+ insurance edge cases resolved pre-handover
Process
How we built it.
Appointment booked in PMS → AI agent sends SMS + email confirmation
Patient reply → AI parses intent (confirm / reschedule / cancel) → updates PMS
Insurance pre-auth document in inbox → AI classifies carrier, plan, status
High-confidence pre-auth → auto-filed against patient record in PMS
Low-confidence document → routed to shared human exception queue
Reminder cadence → AI agent triggers 7-day, 24-hour, and 2-hour touches
Branch dashboard → live view of confirmations, exceptions, and pre-auth status
30-day post-launch tuning → 12+ insurance edge cases resolved pre-handover
Start a project
Front desk drowning in repetitive calls?
Our 4-week Business Automation Package ships one production AI automation for the workflow draining the most hours — fixed scope, weekly demos.
No retainer lock-in · Month-to-month · Full transparency
More work
Up next.
A mid-market freight brokerage replaced the first two hours of every broker's day with an AI agent that parses, matches, and confirms loads automatically.
Atlas Freight Brokerage — Quote-to-Load Automation
From paid-ad dependency and flat AOV to an organic-first growth engine with AI-driven recommendations.