5 hrs+
Saved per doctor per week
Automated collection of pre-appointment information, instead of the doctor chasing it in the room.
Case Study
Multi-channel AI Agent for personalized appointments in Healthcare
The brief was an agent that reads a patient's full record — history, physician notes, active issues, visits, personal details, medications — then gathers what is still missing from the patient, on channels older adults actually use, under HIPAA.
The outcome
Five hours back to the doctor, before the appointment starts
The agent lives in the pre-appointment window. Seniors update concerns by voice or SMS instead of a form. Doctors walk in with a summary already written from the record plus those replies. That is where the hours go, and where engagement moved.
5+ hours per doctor per week and 20%+ engagement are the published outcomes of this rollout, not STCC triage scores and not a lab demo. HIPAA stayed on. Records update after each visit so the next turn is not a blank form.
20%+
Increase in patient engagement
Personalized SMS and voice, available when the patient needed them — not a portal built for younger users.
Client voice
I love the summaries. I think it is amazing, and I am very impressed with this tool.
End user (doctor)
Physician using the pre-appointment agent
About the client
A US healthcare company focused on high-quality, affordable, easy-to-understand plans for seniors. It specializes in Medicare Advantage and uses technology to improve delivery. It operates across multiple US states and serves more than 100,000 members. By the end of that year it employed more than 550 people and is listed on the NASDAQ. The company is not named here.
Vstorm's impact
Vstorm's impact, the TL;DR
- Pre-appointment agent: patients send updates before they see the doctor
- Voice and SMS designed for seniors, instead of a portal or a paper form
- More than five hours saved per doctor per week on information collection
- Patient engagement up more than 20%
- RAG over the full record so questions are patient-specific, under HIPAA
- After each visit, the record updates into a single health history
The challenge
Seniors, many clinics, and a record the doctor should not re-ask
As the organization grew, clinics needed a scalable way to talk to senior patients without burning physician time. They wanted a consulting and engineering partner that could ship agents into existing healthcare systems, not a sidecar chatbot.
The agent had to read the entire record — medical history, physician notes, active issues, visit history, personal details, current medications — then collect what was still missing from the patient, on channels older adults will actually use. It also had to keep building a data foundation for later personalization, all under HIPAA.
How we delivered
TriStorm with doctors still in the loop
The first version was not a black box. A select group of doctors ran a Proof of Value, kept oversight, and only then did automation widen.
Fit the agent to senior workflows
The brief was pre-appointment, HIPAA, and channels seniors already understand. Generic platforms were out: the agent had to sit on clinic-management and physician-management systems already in use.
- HIPAA brief
- Senior-channel map
- Core-platform constraint
Proof of Value with a small doctor cohort
A semi-manual first agent so physicians stayed in control. Exceptions surfaced in real visits. Workflows, model limits, integrations, and change management were fixed before the rollout grew. Doctors noticed the time it saved; that is what unlocked scale.
- Doctor-tested agent
- Exception log
- Clinic-system integration
SMS, voice, then a living record
Multi-channel outreach. RAG over the complete history so questions are not generic. After each appointment the record updates itself. Engagement rose more than 20%; each doctor gained five-plus hours a week.
- SMS and voice
- RAG over the chart
- Post-visit record write-back
How it works
The chart, then a voice call, then a summary
The agent is for the pre-appointment phase. It reads the record, uses retrieval-augmented generation to ask questions that belong to that patient, and reaches the member by SMS or voice — not a form built for a younger portal user. Voice is the more natural path for many older adults. The doctor gets a summary. After the visit, the new facts land back in one health history.
Patient record History, notes, meds, visits
↓
RAG Patient-specific questions, not a generic script
↓
SMS or voice Channels designed for seniors
↓
Doctor summary Updates and concerns before the room
↓
Record write-back Centralized history after the appointment
Vstorm × undisclosed US Medicare Advantage payer — pre-appointment agent
Off-the-shelf platforms were rejected so the agent could follow this organization's workflows and talk to clinic and physician systems without a second round of manual entry. HIPAA applied to the whole path.
Results
What changed before the appointment
Each doctor saved more than five hours a week. Patient engagement rose by more than 20%. The hours come from automated collection and a data foundation for later personalization, not from a generic questionnaire. Compliance held. Patients and care teams both used it.
Work with us
Ready to see how agentic AI transforms pre-appointment workflows?
Meet directly with our founders and PhD AI engineers. We will walk through real implementations from 30+ agentic projects and the practical steps to integrate them into your workflows.