Case study

AI Pharmacy Communication Assistant

A Twilio + LLM assistant that handles refill confirmations, delivery updates, and patient data corrections — HIPAA-aware from day one.

Section 01 · The problem

What was actually breaking before we touched it.

01

Staff was spending most of every day on the same handful of conversations: refill confirmations, delivery times, and address corrections.

02

Patient contact and address data drifted out of date constantly — and nobody had time to clean it.

03

Generic AI tools weren't an option: anything customer-facing needed to operate inside HIPAA-aware boundaries.

The shape of what we built
data pipeline · live
01
Inbound
SMS · voice
02
Understand
LLM intent
03
Confirm
Patient reply
04
Update
PMS write-back
05
Escalate
Human-in-loop
+ resilient retries+ proxy rotation+ scheduled refresh+ change detection
Section 02 · The system

What we built and how it works.

01

Two-way SMS and voice flows handling delivery confirmations, refill scheduling, and address corrections.

02

Structured extraction from patient replies that updates the pharmacy database automatically when confirmed.

03

Hard escalation paths for anything outside the assistant's safe lane — clinical questions, complaints, or ambiguous responses.

04

HIPAA-aware logging, access control, and audit trails on every conversation.

Section 03 · The impact

What changed for the operator.

01

The repetitive top of the staff's queue effectively disappeared.

02

Patient contact and address records started staying current automatically.

03

Patients got faster answers without losing the option to reach a human.

Tell us what's broken.
We'll tell you what to build.

A 30-minute call. No pitch deck. We map the workflow you're trying to fix and tell you straight whether we're the right team for it.