AI-Automated Care Management

FlorencebyNightingaleMD

The AI That Does Everything Except the Nursing

"Your team — everywhere, at once."

Florence automates 47 care management workflows — scheduling, outreach, follow-up, referral tracking — giving nurses and care managers the clinical calm to focus on what only they can do.

Built from human factors research. Designed with nurses, for nurses.

47
Workflows Automated
↑90%
Top-of-License Time
24/7
AI Voice Engagement

How Florence Works

Detect the right member, engage in a human way, then act safely with clinical guardrails.

Detect

Ingests clinical intelligence from EHR, HIE, ADT, claims, labs, and more.

Engage

Reaches members by voice and SMS in their preferred language.

Act Safely

Executes with guardrails: RN/MD escalation, EHR write-back, and a full audit trail.

HIPAA BAA Available SOC 2 Type II Full Audit Trail CMS Conversational AI Early Adopter '25 ↗
From Research to Product

We Didn't Start With the Technology.
We Started With the Nurses.

Before writing a line of code, we conducted a human factors study of care management — cognitive task analysis, time-motion studies, nurse shadowing, and subject matter expert workshops. We mapped the full landscape of care coordination work.

322
Workflows mapped
across 18 domains
40%
Automatable without
clinical judgment
47
Workflows Florence
automates today
60→90%
Top-of-License
clinical time

We found exactly what nurses needed — and built it. Florence is the product that fulfills the spec our research produced. Every feature maps to a real workflow, a real time sink, a real failure mode we observed.

Read the Full Research Blueprint →

Trusted by Value-Based Care Leaders

ACO Network Partners
Medicare Advantage Plans
FQHC Systems

"Florence has transformed how our care navigators manage their panels — we've seen a 3x improvement in engagement rates." — VP of Population Health

Explore Florence for Your Role

Florence serves every stakeholder—providers, nurses, plans, and patients. Choose your perspective.

For Nurses & Care Teams

You became a nurse to care for people.
Florence makes sure you can.

Florence handles the repetitive outreach, documentation, and scheduling — so you can spend your time where it actually matters: with patients who need your clinical judgment, your empathy, and your expertise.

24712

Prioritized Worklists, Not Endless Calls

Start each day with a sorted list of patients who need you — not a queue of voicemails. Florence completes initial outreach and surfaces only the patients requiring clinical attention.

F

Documentation That Writes Itself

Every call Florence makes generates structured notes in your EHR — SOAP format, ICD codes suggested, care plan updates queued for your review. Less charting, more caring.

Warm Handoffs, Never Cold Drops

When a patient needs you, Florence doesn't just transfer — it briefs you first. Full context, red flags highlighted, patient mood assessed. You walk in prepared, not cold.

"We didn't build Florence to take your job. We built it because your job is too important to bury in paperwork."

See how Florence gives nurses back their time.

Schedule a 30-Minute Workflow Review →
A Day in the Life

Dana, RN Care Manager — ACO Network

7:45 AM — Before her first cup of coffee is finished

Dana opens her EHR and sees a streamlined worklist: just 8 high-need patients from 47 recent discharges who require her direct attention.

Florence has already handled the other 39:

  • Placed and transcribed automated check-in calls for all 39 patients using structured voice assessments
  • Cross-referenced medication lists with discharge orders, flagging 3 discrepancies for pharmacy review
  • Confirmed PCP follow-up appointments, identifying 5 patients needing transportation assistance
  • Triaged symptoms — escalated 1 chest-tightness case with transcript and call audio for Dana's immediate review
  • Documented every interaction directly in the EHR, pre-populating Dana's notes

By 10 AM, Dana has resolved the urgent case and spoken to every high-acuity patient. Half her documentation is already complete — auto-filled by Florence. She's practicing at the top of her license — not chasing voicemails.

Three Failure Modes Draining Your Care Team

We studied how care management actually works — shadowing nurses, running time-motion analysis, mapping 322 workflows. These are the three patterns that break every team.

Phone Tag Math

A nurse with 200 patients makes 8 call attempts per engagement. That's 1,600 dials per cycle — most of which end in voicemail. The math never works.

Florence takes the first 1,600 dials off your team's plate — automating outreach, scheduling callbacks, and triaging who actually needs a nurse on the line. Over 60% of phone time goes back to clinical work.

The Referral Black Hole

Referrals go out and vanish. No one tracks whether the patient saw the specialist, whether charts came back, or whether the loop closed. Patients fall through silently.

Florence tracks every referral from the moment it's sent — confirming specialist appointments, retrieving external charts, and flagging open loops until the care circle closes.

The Equity Penalty

When there aren't enough hours, lower-acuity patients get deprioritized. But "lower-acuity" doesn't mean "lower-need." Entire populations go unmanaged — not by choice, but by capacity.

Florence reaches the full panel — not just the top decile. Chronic condition check-ins, proactive wellness outreach, and early-warning escalation happen at scale, so no patient is managed only by default.

These aren't technology problems. They're human factors problems — and they require a human factors solution.

Ready to see how Florence handles these workflows?

Schedule a 30-Minute Workflow Review →
47 Workflows Across 8 Domains

The Automation Surface

Of 322 care management workflows we mapped, 47 don't require clinical judgment. Florence automates these — so your team's time goes to patient care, not logistics.

Patient Outreach

12 workflows

65% automatable

Scheduling

8 workflows

75% automatable

Referral Mgmt

7 workflows

55% automatable

Documentation

6 workflows

45% automatable

Med Adherence

5 workflows

50% automatable

Care Transitions

4 workflows

35% automatable

Quality Measures

3 workflows

40% automatable

Chronic Care

2 workflows

30% automatable

The principle: Florence automates only what doesn't require clinical judgment. Nurses keep the decisions. Florence handles the logistics.

Measures We Move

When nurses spend 90% of their time on clinical work instead of 60%, quality metrics move with them.

HCC Recapture Rate

Baseline 45%
+20%
Goal 85%

AWV Completion Rate

Baseline 30%
+25%
Goal 75%

HEDIS Quality Measures

Baseline 40%
+30%
Goal 85%

ED Diversion Rate

10%
+20%
Goal 45%

TCM Follow-Up (48hr)

Baseline 35%
+35%
Goal 90%

Patient Engagement Rate

20%
+40%
Goal 80%
A Day in the Life

Sarah's Tuesday With Florence

Sarah is a care manager with 200 patients. Before Florence, most of her day was logistics. Here's what changed.

7:15 AM

Sarah opens her dashboard. Florence has already completed overnight outreach to 34 patients — medication reminders, appointment confirmations, and post-discharge check-ins. Twelve responded. Three need follow-up.

Florence handled: 34 outreach calls, response triage, priority sorting

9:00 AM

Mr. Rivera's GFR is trending down. Florence flagged it last night — pulled his latest labs, checked his medication list against his diagnoses, and surfaced a potential drug interaction. Sarah spends 15 minutes on clinical judgment, not 45 minutes gathering data.

Florence handled: Lab monitoring, chart aggregation, interaction check, alert prioritization

11:30 AM

A referral from three weeks ago never closed. Florence tracked the specialist appointment, confirmed attendance, retrieved the consult note, and routed it to the PCP — before Sarah even knew there was a problem.

Florence handled: Referral tracking, chart retrieval, loop closure, PCP notification

3:00 PM

Sarah reviews her end-of-day summary. She personally engaged with 18 patients today — all clinical conversations. Florence handled the other 182. No one fell through the cracks. Her Top-of-License time: 91%.

The result: 200 patients managed. 18 by Sarah. 182 by Florence. Zero missed.

The 5-Phase Care Worklist

End-to-End AI-Automated Care Management

From signal detection to continuous surveillance — Florence, your AI nurse extender, orchestrates every step. Click any phase below to expand its capabilities.

Gap Journey: Detected Surfaced Addressed Closed Maintained
1
Trigger
Signal detection and workflow initiation
Signal DetectionWorkflow InitiationLiving Document

Florence AI Capabilities

Signal Detection from PopHealth
Automatically detects care opportunities from your Care Orchestration Platform
Workflow Initiation
Signal Detection — Automatic workflow initiation
Living Document Created
VBC Process Activated with continuous visibility thread

PopHealth Signals Detected

AWV (Annual Wellness Visit)
Preventive care opportunity identified
ADT (Admission/Discharge/Transfer)
Transition of care event detected
Care Gap / HCC/RAF Opportunity
Quality measure or risk adjustment need identified
2
Pre-Encounter
Patient engagement and gap surfacing
Multi-Channel20+ LanguagesGap Pre-ID

Florence AI Capabilities

Multi-Channel Outreach
SMS • Voice • Chat — reaching patients where they are
20+ Languages
Native engagement in the patient's preferred language
Gap Pre-Identification
Surfacing gaps before the visit so the care team is prepared
Patient Activation
Education and preparation for the upcoming encounter

Human Actions

Patient Advocacy
Florence acts as patient advocate, coordinating care
Medication Reconciliation
Conversations to verify current medications
Transportation & Caregiver
Coordination and engagement support
3
Encounter
PCP Visit with Florence Copilot
Care GapsDiagnosisScreeningSDOH

Florence Copilot Domains

Care Gaps
Comprehensive Assessment Completion — Gap → Addressed
Diagnosis
HCC Capture • RAF Optimization • Coding Accuracy
Screening
HEDIS Measures • Preventive Care • Quality Gaps
SDOH
Social Determinants • Community Resources • Referrals
ZERO DOUBLE-ENTRY
Work in Copilot → Florence Auto-Files to EHR: Progress Notes, Medications, Problem List, Quality Attestations

Care Team Actions

Physician
Fills structured assessment in Florence Copilot — Florence nudges for completion
Navigator
Supervises remotely with real-time visibility — No need to be in room
Patient
Center of care, active participant — Florence as advocate, gaps addressed in real-time
4
Post-Encounter
Automatic filing and attestation
Auto-FilingAttestationPopHealth Sync

Florence AI Capabilities

Automatic Filing
Automated Write-Back to all EHR sections — eliminates copying/pasting
Attestation Workflows
Quality measure closure with validated ICD-10 and CPT-II codes
PopHealth Sync
All activity reported — Bidirectional synchronization

Human Actions

Chart Review
EHR Reconciliation — Document Viewer without toggling systems
Quality Attestation
HEDIS measure closure with documentation
Referral Initiation
Loop tracking started, care plan finalized
5
Care Management
Continuous surveillance and relationship
LongitudinalProactive OutreachGap Surveillance

Florence AI Capabilities

Longitudinal Relationship
Continuous patient engagement over time
Proactive Outreach
Early intervention before issues escalate
Continuous Gap Surveillance
New needs detected → Loops back to TRIGGER — The continuous loop that ensures nothing falls through the cracks

Human Actions

Adherence Tracking
Care plan compliance monitoring
Referral Loop Closure
Monitoring completion of specialist referrals
Medication Adherence
Ongoing checks — Patient maintained in care plan

Three-Way Call Capability

Florence bridges patients, caregivers, and care navigators in real-time — 400,000+ contacts per year vs. 4 traditional visits

Patient

Receives the call, shares symptoms, concerns, and daily status

Caregiver

Joins when needed for context, medication verification, or support

Florence AI

Orchestrating the Conversation

Condition-Specific Clinical Protocols

Deep clinical intelligence across the most common chronic conditions in value-based care

20+ Languages
View Full Protocol Library →

Click any category to explore how Florence manages each condition

Cardiovascular

5 conditions
  • Congestive Heart Failure (CHF)

    Florence automates daily weight and symptom monitoring per ACC/AHA guidelines, escalating to the care team when weight gain exceeds 3 lbs in 24 hours or 5 lbs in a week. Ensures medication adherence for beta-blockers, ACE inhibitors, and diuretics while tracking fluid restriction compliance.

  • Hypertension

    Florence conducts protocol-driven blood pressure monitoring, medication adherence checks, and lifestyle coaching on diet and exercise. Automatically escalates when BP readings fall outside patient-specific target ranges, reducing ED visits and stroke risk.

  • Atrial Fibrillation

    Florence monitors anticoagulation adherence (warfarin, DOACs), tracks heart rate control, and screens for stroke symptoms using validated protocols. Coordinates INR lab scheduling and escalates when patients report palpitations, dizziness, or medication side effects.

  • Coronary Artery Disease

    Florence supports cardiac rehabilitation engagement, medication reconciliation (statins, antiplatelets), and lifestyle modification counseling. Tracks chest pain episodes and activity tolerance, escalating when anginal symptoms change or worsen.

  • Post-MI Care

    Florence guides patients through the critical 30-day post-discharge window with daily check-ins on chest pain, shortness of breath, and medication adherence. Automates cardiac rehab referral follow-up and ensures completion of post-MI quality measures.

Respiratory

4 conditions
  • COPD

    Florence monitors symptom exacerbations using COPD Assessment Test (CAT) scores, tracks rescue inhaler usage, and ensures maintenance therapy adherence. Escalates when patients report increased dyspnea, sputum production, or fever—preventing costly hospitalizations.

  • Asthma

    Florence conducts asthma control assessments per NHLBI guidelines, tracks peak flow readings, and monitors controller medication adherence. Identifies trigger exposures and escalates when patients require frequent rescue inhaler use or nocturnal symptoms increase.

  • Pulmonary Fibrosis

    Florence monitors oxygen saturation, dyspnea scores, and functional status while coordinating pulmonary rehabilitation and antifibrotic medication adherence. Escalates when O2 sats drop below prescribed thresholds or patients report worsening breathlessness.

  • Sleep Apnea

    Florence ensures CPAP adherence by tracking nightly usage hours, troubleshooting mask fit issues, and monitoring daytime fatigue symptoms. Coordinates DME vendor communication and escalates when compliance falls below the 4-hour nightly threshold.

A Phased Approach to Autonomy

Start with oversight. Graduate to scale.

We meet you where you are. Begin with full care-team partnership and progressively expand your nurses’ reach as your team builds trust and confidence.

What Your Patients Experience

Florence becomes your patients' active health advocate — closing gaps, coordinating care, and following through.

Protocol-driven outreach: Calls timed to their treatment plan
Multilingual support: 20+ languages, culturally sensitive
Consistent follow-through: Never misses a scheduled check-in
Warm handoffs: Seamless escalation to your nurse or physician when needed
Phase 1

Co-Pilot

Human in the Loop

  • CM listens to every call
  • Reviews all documentation
  • 100% Supervision
Duration: 1-3 months
Start with Co-Pilot
Phase 2

Supervised

Human on the Loop

  • Florence acts autonomously
  • CM monitors dashboard
  • 1 CM : 10-20 Calls
Duration: 3-6 months
Explore Supervised
Phase 3

Full-Scale Partnership

Your Care Team, Amplified

  • Florence handles routine outreach end-to-end
  • Escalates to your nurse or physician on risk
  • 24/7 Scale
Ongoing
Expand Your Reach
Listen

Hear Florence on a Real Call

A short clip from a post-discharge follow-up, with Florence confirming meds, symptoms, and next steps before looping in a nurse when needed.

Watch Full Demo

Transcript Excerpt (Anonymized)

Florence: Hi, I'm calling to check in after your recent discharge. Are you in a safe place to talk for two minutes?
Member: Yes.
Florence: Great. Let's confirm your medications and symptoms, then I'll help schedule your follow-up if needed.

Transitional Care Management — Without Burning Out Your Nurses

Before Florence

  • Nurse manually calls every discharged patient within 48 hours
  • 60% of calls go to voicemail — nurse calls back repeatedly
  • Documentation takes 15+ minutes per patient
  • High-acuity patients wait while nurse chases routine follow-ups

With Florence

  • Florence completes Day 1 outreach for every discharge automatically
  • Nurse gets a prioritized list: who needs a human call now
  • Structured notes in EHR before nurse picks up the chart
  • Nurse time focused on clinical judgment, not phone tag

Early programs show meaningful reductions in 30-day readmissions and significant time returned to nurse care managers per TCM episode.

How Florence Communicates

Florence's Five-Pillar Therapeutic Persona

Florence delivers chronic care management that feels human—combining comprehensive data access with empathetic conversation to transform care management.

1

Psychological Safety

Creates a judgment-free space where patients feel secure sharing concerns

  • Non-judgmental language patterns
  • Validates patient experiences
  • "Tell me more about that" prompting
2

Warmth

Conveys genuine care through tone, pacing, and personalized details

  • Remembers previous conversations
  • Adjusts pacing to patient needs
  • Uses patient's preferred name
3

Competence

Demonstrates clinical knowledge while staying within scope

  • Accurate medical terminology
  • Clear scope boundaries
  • "Let me connect you with..." when appropriate
4

Inquisitiveness

Asks thoughtful questions to understand the patient's experience

  • Open-ended questioning
  • Explores the full patient context
  • "How has this affected you?" focus
5

Unfailingly Patient

Gives time for reflection and never rushes important topics

  • No time pressure cues
  • Pauses for patient processing
  • Repeats info without frustration
Florence in Action

Your AI Nurse Extender, Embedded in the EHR

Florence brings clinical intelligence directly into your workflow. No portals, no toggling, no delays.

healthchart-ehr.com/patient/jane-doe
Chart
Labs
Notes
Vitals
Schedule
JD

Jane Doe

75 y/o Female • MRN: 847291 • DOB: 03/15/1950

CHF, T2DM, HTN, CKD Stage 4
High Risk Discharged

Active Problems

I50.9 Heart Failure, unspecified
E11.9 Type 2 Diabetes Mellitus
I10 Essential Hypertension
N18.4 CKD Stage 4

Upcoming Appointments

15 Feb
Cardiology Follow-up
Dr. Michael Brooks • 2:30 PM
22 Feb
Nephrology Consult
Dr. Lisa Wong • 11:00 AM

Recent Lab Results

eGFR
29 (Low)
Creatinine
2.1 (High)
HbA1c
7.2%
BNP
450 pg/mL
Potassium
4.2 mEq/L
EHR Copilot
Active
Facesheet
Transcript
Documentation
Log

Florence Summary & Agenda

Jane was discharged from Memorial Hospital 3 days ago following CHF exacerbation. Recent labs indicate CKD progression (eGFR 29). TCM follow-up required within 7 days per CMS guidelines.

Action Items:

  1. Complete TCM Day 1 outreach call
  2. Discuss recent CHF discharge and medication changes
  3. Schedule PCP follow-up within 7 days
  4. Address suspected CKD Stage 4 (eGFR 29)
ADT Alert Discharge

Discharged from Memorial Hospital - 3 days ago

Reason: CHF Exacerbation • LOS: 4 days

1 critical alert needs attention
3 care gaps need attention

Gap Journey

40% Closed
5 Detected
1 Surfaced
1 In Progress
1 Addressed
2 Closed

CKD Stage 4 Detected

eGFR 29 indicates progression • ICD-10: N18.4

TCM Follow-up Due

Post-discharge visit within 7 days required

AWV Eligible

Annual Wellness Visit due March 2026

Clinical Note

Auto-Generated

Documentation will appear here after completing a workflow.

CKD Stage 4 Filed to EHR

ICD-10 N18.4 added to problem list

2 min ago

ADT Alert Received

Discharge from Memorial Hospital

3 days ago

Lab Results Analyzed

eGFR 29 flagged as critical

3 days ago
Supervised Autonomy Dashboard
1 Care Manager : 12 Active Calls
12
Active Calls
2
Needs Review
47
Completed Today

We don't replace nurses. We protect them.

Florence never replaces a nurse's clinical judgment. Every escalation, every care decision, every patient concern that needs a human — it goes to your team.

We reduce burnout, not staff. Florence handles the outreach volume so your nurses can work at the top of their license — not buried in phone tag and paperwork.

We honor the people who chose this profession. Every feature we build starts with one question: does this give a nurse more time to do what only a nurse can do?

Florence's Escalation Protocol: Safety Guardrails Built In

Every scenario has a defined response—nothing falls through the cracks

Click any row to see detailed safety protocols and audit trails.

TriggerResponseOutcome
Medical Emergency

High-risk keywords: bleeding, chest pain, falls, unresponsive

Alert RN immediately + warm transfer to clinical team

Immediate escalation
Detailed Safety Protocol

Trigger Detection: NLP model trained on 10,000+ emergency transcripts flags critical keywords in real-time. Confidence threshold: 95%.

Response Workflow:

  • RN receives SMS + app notification within 10 seconds
  • Florence keeps patient on line while connecting RN
  • Warm handoff includes full conversation context
  • If RN doesn't respond in 60s, escalates to on-call MD

Audit Trail: Full call recording + transcript stored for 7 years (HIPAA compliant). Incident report auto-generated to Quality team within 15 minutes.

Suicidal Ideation

PHQ-9 Q9 positive, self-harm language

PHQ-9 screening protocol + RN escalation + safety plan

Non-interruptible handoff
Detailed Safety Protocol

Trigger Detection: Any positive response to PHQ-9 Question 9 ("thoughts of self-harm") or passive language indicators ("better off without me").

Response Workflow:

  • Florence immediately administers Columbia Suicide Severity Rating Scale (C-SSRS)
  • RN joins call within 2 minutes (median response: 47 seconds)
  • Safety plan documented: support contacts, crisis hotline, emergency services if needed
  • Follow-up scheduled within 24 hours, tracked until completion

Compliance: Meets Joint Commission sentinel event reporting requirements. Behavioral health specialist notified same-day.

Elder Abuse / Neglect

Signs of abuse, financial exploitation, neglect indicators

Mandatory reporting protocol + social work referral

Clinician + SW review
Detailed Safety Protocol

Trigger Detection: Multi-factor model assesses: unexplained injuries, caregiver control of conversation, financial concerns, hygiene/nutrition issues, isolation from family.

Response Workflow:

  • Case flagged to Social Work team within 1 hour
  • SW conducts comprehensive assessment within 48 hours
  • Mandatory state reporting filed per jurisdictional requirements (Adult Protective Services)
  • Care team coordinates with legal/compliance for documentation

Legal Compliance: Adheres to state-specific mandatory reporting laws. Full chain-of-custody documentation for potential legal proceedings.

Behavioral Crisis

Aggression, wandering/safety risk, severe agitation

Escalation flag + full transcript to care navigator

Care manager review
Detailed Safety Protocol

Trigger Detection: Vocal stress analysis + behavioral pattern recognition. Indicators: raised voice, rapid speech, disorientation, expressed intent to leave safe environment.

Response Workflow:

  • Florence uses de-escalation techniques (validation, redirection, empathy)
  • If unsuccessful after 2 minutes, care manager joins call
  • Full transcript + sentiment analysis delivered to CM dashboard
  • CM evaluates for: medication review, environmental safety, psychiatric consult

Follow-up: Care plan updated. Follow-up call scheduled within 24-48 hours to assess resolution.

Confusion Detected

Patient unable to answer, disoriented, impaired

Request caregiver join call immediately

Three-way call
Detailed Safety Protocol

Trigger Detection: Inability to recall basic info (name, date, address), repetitive questions, tangential responses, long pauses, or explicit statements of confusion.

Response Workflow:

  • Florence asks: "May I call [caregiver name] to join our conversation?"
  • Three-way call initiated using emergency contact on file
  • If caregiver unavailable, Florence documents observations and notifies care team
  • Care team evaluates for: cognitive decline, medication side effects, acute delirium

Clinical Follow-up: MD or NP review flagged within 24 hours. May trigger cognitive assessment (MMSE/MoCA) at next visit.

Routine Call

No flags, standard check-in

Document findings + schedule follow-up

Auto-note to EHR
Detailed Safety Protocol

Standard Workflow: Florence completes protocol-driven assessment (symptom check, medication adherence, vital signs if applicable, appointment reminders).

Documentation:

  • Structured note auto-filed to EHR within 2 minutes of call completion
  • Includes: call duration, topics covered, patient-reported outcomes, action items
  • Care gaps closed (e.g., mammogram scheduled) reflected in quality dashboards
  • Next follow-up auto-scheduled per protocol (e.g., CHF weekly, diabetes biweekly)

Quality Assurance: 10% of routine calls randomly selected for RN quality review to maintain calibration and identify improvement opportunities.

Florence Works Where Your Team Already Works

No new portals. No toggling between systems. Florence reads from and writes back to the EHR your nurses already live in.

EHR Systems

Epic
athenahealth
eClinicalWorks
NextGen
Allscripts
MEDITECH
Oracle Health
Greenway
ModMed
AdvancedMD
DrChrono
Kareo

Population Health Platforms

Arcadia
Innovaccer
Health Catalyst
Lightbeam
Azara
Epic Healthy Planet
Salesforce Health

Clinical Data Sources

EHR/CCDA
HIE Networks
Hospital ADT
Lab Networks
Pharmacy
750K+
Clinicians Connected
147K+
Care Sites Linked
FAQ

Questions Your Team Is Already Asking

Straight answers, no scavenger hunt.

How does Florence integrate with our EHR and data sources?
Florence connects via standards-based interfaces (FHIR/HL7 where available) and secure APIs. Typical integrations cover member context, outreach triggers, tasking, and documentation outputs. We support EHR, HIE, ADT feeds, claims, and lab data — and our team handles the integration work so your IT burden is minimal.
What happens if a member is suicidal or in a medical emergency?
Florence follows hard-stop escalation rules for crisis scenarios. Suicidal ideation, chest pain, and other emergencies trigger immediate routing to your defined clinical pathway — RN triage, 911, or emergency resources — with full auditable logging. Florence is never the last line of defense.
How do you prevent hallucinations in clinical workflows?
Florence uses protocol-driven workflows with confidence thresholds, not open-ended generation. When the model is uncertain, the interaction routes to a human. All outputs are designed to be reviewable, traceable, and auditable — never a black box.
How fast can we get a pilot live?
Most pilots start with one prioritized use case, clear success metrics, and a 90-day implementation plan. We handle the heavy lifting on integration and protocol configuration. Typical time-to-live is weeks, not quarters.
What data does Florence store, and for how long?
Data handling is governed by your organization's security requirements and retention policies. We provide encryption at rest and in transit, role-based access controls, and complete audit logs to support HIPAA compliance for PHI workflows.
How does Florence escalate to a human?
Escalation triggers can be clinical (symptoms, risk signals, confidence thresholds) or operational (member requests, uncertainty). When triggered, Florence routes the interaction to the right role — RN for clinical judgment, MD for high-risk — and packages the full conversation context for fast decision-making.

What Florence is NOT: Florence is not a phone tree, not a generic chatbot, and not a debt collector. It's a clinical-grade voice assistant built for value-based care.

Research Paper

The AI-Automated Care Management Blueprint

26 pages of human factors research — 322 workflows mapped, three failure modes identified, and the specification that became Florence. The complete case for why care management automation requires a human factors approach.

We'll send the PDF and nothing else. No spam.

Ready to Give Your Care Team Back Their Time?

See how Florence automates 47 care management workflows — so nurses stay at the top of their license.

30-minute workflow review · No commitment · HIPAA-compliant