Licensed Operator · Platform Product Leader · 20+ yrs Colorado · Fort Collins · Open to remote

The care delivery layer
is where AI has to earn it.

I'm Sarah Brock. I have 20+ years in healthcare technology, split across two chapters that don't usually exist in the same person: 8 years as a licensed operator running post-acute facilities, and 12+ years as a product leader building the software those facilities depend on. The first chapter is why the second one works. These prototypes show how I think about AI in healthcare — how I identify real clinical and operational problems, scope AI-powered solutions, and build working examples to validate the concept.

See the three prototypes Email Sarah LinkedIn
These prototypes were built using Claude and AWS Bedrock — working examples, not slide decks.
4 min
Prior auth package
generated by AI
vs 2+ hours manual
31%
Industry avg prior auth
denial rate
problem CarePathIQ is built to address
1,600+
Providers served at
HOPE regulatory launch
only vendor ready at CMS deadline
67%
CHF readmission
prediction accuracy
3-signal simultaneous detection
Platform thesis

Healthcare doesn't lack data.
It lacks systems that learn from it.

Three structural failures repeat across every care setting I've worked in — as an operator and as a product leader. Clinical knowledge disappears between shifts. Episode-level decisions happen without outcome feedback. Regulatory data sits disconnected from the care delivery layer where it's most needed.

AI applied to documentation and chat interfaces doesn't fix those failures. The opportunity is a care intelligence platform that connects operational knowledge, clinical decisions, and outcome data across the full episode of care. The prototypes below explore what three surfaces of that platform could look like — built with Claude and AWS Bedrock, grounded in workflows I've operated inside.

01 — PRESERVE
ShiftIQ
Operational knowledge preservation — capturing what nurses tell each other at shift change that never makes it into the chart.
02 — PLAN
CarePathIQ
Episode pathway intelligence — routing each patient to the right post-acute path, with prior auth generated in minutes.
03 — MONITOR
ContinuIQ
30-day episode intelligence — watching weight, therapy engagement, and family contact simultaneously to catch deterioration early.
DESIGN CONSTRAINT — CONTEXT ARCHITECTURE
applies to all three surfaces
Clinical AI can't run out of context.
Full stop.

Context windows are finite. Clinical workflows are not. When session pressure causes a clinician to rush or skip clarifying back-and-forth, incomplete information enters the record and corrupts every downstream decision that depends on it. This isn't a UX problem — it's a care quality failure that propagates across all three platform surfaces.

CarePathIQ · Prior Auth
WRONG AUTH RISK
The failure:

Mid-review, CHF + COPD comorbidity interaction surfaces. That combination requires shorter therapy sessions across more days — a different authorization structure entirely. Under context pressure, you can't reintroduce that interaction and rerun the logic cleanly.

Consequence:

Patient approved for fewer SNF days than clinically needed. Discharged before functional baseline. Readmission follows.

ShiftIQ · Handoff · 07:00
DATA GAP
The agent asks:

"No therapy updates for Bob across 3 shifts — unusual for his care plan. Did therapy not occur, or does it need to be documented? Marking did-not-occur vs. undocumented changes his episode metrics, readmission risk, and current authorization coverage."

Why this matters:

Proactive gap detection before the handoff closes. The agent notices what's missing and surfaces it while there's still context to act on it.

These aren't isolated failures. A missing therapy note in ShiftIQ means ContinuIQ monitors against a flawed baseline. A wrong auth from CarePathIQ becomes a readmission that ContinuIQ flags two weeks later. Context failure in one surface propagates through all three.

ARCHITECTURAL REQUIREMENTS — NOT FEATURES TO ADD LATER
01
Session Continuity
Workflows persist across the full clinical interaction, not until a token budget runs out.
02
Graceful Degradation
Context limits compress intelligently. Ceiling pressure never reaches the clinician.
03
Proactive Gap Detection
Agent surfaces missing information before the workflow closes, not after.
04
Tiered Memory
Persistent patient context lives separately from active session context. The model knows which is which.
Prototypes

Three surfaces.
One platform thesis.

Each prototype addresses a distinct failure point in post-acute care. Click any card to explore the full interactive prototype — built with Claude and AWS Bedrock, fictional patient data throughout.

Preserve
Plan
Monitor
shiftiq · shift handoff intelligence · Wing 2 North
click to explore ↗
Night → Day · 07:00 3 priority alerts ✦ AI extracted
201A · Mary J. PAIN
Says 4/10. Treat as 7. Watch grimace, not the number.
210A · Bob C. APHASIA
Write your name on the whiteboard before you do anything else. Every single visit.
203B · George W. FAMILY
Son hasn't visited in 9 days. Asks about him each morning. Flag for social work.
✦ Extracted from 11 shift notes · Voice · Photo · Type
01 — SHIFT INTELLIGENCE · HUMAN KNOWLEDGE TRANSFER
ShiftIQ
At shift change, nurses tell each other things that never make it into the chart. Mary says 4/10 — treat her as 7. Bob needs your name on the whiteboard before anything else. ShiftIQ captures that knowledge and puts it in the next clinician's hands before they walk into the room. Click to see a full Night→Day handoff.
SBAR
AI-generated
handoff format
RN+CNA
Role-specific
views
5 pt
Full patient
profiles
Interactive · fictional patient data
Open prototype
carepathiq · pathway comparison · prior authorization
click to explore ↗
DRG 470 — same procedure, same OR ✦ different paths
Margaret S. SNF first
COPD · MCI · FIM 68
14 meds · High complexity
Robert K. Home first
Spouse caregiver · FIM 104
5 meds · Low complexity
4 min prior auth generated✦ AI
94% approval probability94%
847 similar patients analyzedlive model
Concurrent reviews auto-scheduled
02 — CARE TRANSITIONS · EPISODE INTELLIGENCE · PRIOR AUTH
CarePathIQ
Two patients leave the same OR on the same day with the same billing code. Their post-acute paths should look nothing alike. CarePathIQ reads the full clinical picture — comorbidities, caregiver support, cognitive load, social determinants — and routes each patient differently, then generates the prior auth package in 4 minutes. Click to see Margaret and Robert's side-by-side comparison.
94%
Avg approval
probability
4 min
Auth package
generated
$35B
Prior auth
market cost
Interactive · fictional patient data
Open prototype
continuiq · 30-day episode intelligence · 2-West wing
click to explore ↗
2-West Wing · 17 residents · Day 14 3 on watch
Eleanor M. · Rm 204 WATCH
weight ↓3.2lb · PT 34% · family 6 days
George W. · Rm 203 CHF WATCH
4-day weight gap · missed Monday weigh-in
Alice P. · Rm 201 on track
✦ AI Eleanor: weight ↓ + PT engagement ↓ + family visits ↓ simultaneously = CHF readmission pattern. 67% historical accuracy. Recommend physician review today.
03 — LONGITUDINAL INTELLIGENCE · 30-DAY EPISODE
ContinuIQ
Clinical data transfers between care settings. The pattern that predicts readmission doesn't. ContinuIQ watches weight, therapy engagement, and family contact simultaneously — and caught Eleanor's CHF risk 4 days before it would have become a hospital transfer. Click to see the 2-West wing census and the alert that triggered.
67%
CHF readmit
prediction accuracy
30-day
Episode
monitoring
$20K
Avg readmit
cost prevented
Interactive · fictional patient data
Open prototype
✦ PLATFORM THESIS These aren't three separate tools — they're three surfaces of a single care intelligence platform. CarePathIQ plans the episode at discharge. ContinuIQ monitors it across every care setting. ShiftIQ preserves the human knowledge that makes the clinical difference at each shift. The insight that connects them: the most important care decisions happen in the gaps between systems, and that's exactly where AI earns its place.
Proof of work

AI built and shipped, not just described.

The three prototypes above show clinical product thinking. This shows the production-track AI work behind them — a working churn prediction model built across enterprise post-acute platforms, validated against CareCore EHR behavioral data, and approved for the roadmap.

Why it matters

The numbers behind the problem.

These aren't prototype metrics — they're the market realities these tools are designed to address.

$300B
Administrative waste, annually
Prior auth alone accounts for $35B+ in payer-provider friction every year in the US healthcare system
97%
Reduction in prior auth time
From 2+ hours of manual research and assembly to 4 minutes with AI-matched clinical documentation
$4–8M
Readmissions prevented
10% rate reduction across a 20-SNF network at $20K average cost per readmission event annually
$7.7M
Revenue at risk surfaced by AI
AAP prototype identified at-risk enterprise post-acute platform accounts 6–12 months before churn — 94% accuracy, 6.4:1 projected ROI, executive buy-in secured
About Sarah

Sarah Brock

Principal PM Licensed NHA · Colorado 20+ yrs healthcare tech Fort Collins, CO

My 20+ years in healthcare technology break into two distinct chapters, and the combination is genuinely rare. The first 8 years I was an operator — licensed to run skilled nursing facilities, overseeing all health information technology across every care setting Columbine operated, doing the hiring, the state licensing, the midnight calls when a resident went to the ER. The next 12+ years I moved to the vendor side, building the software I used to buy and configure and complain about. That sequence matters. When I write a product spec, I know what it feels like when the software doesn't match how care actually happens at shift change. When I talk to a clinical user, I'm not translating — I've been them. That's the foundation these prototypes are built on.

🏥 Licensed NHA · Colorado — opened two skilled nursing facilities from scratch, including site selection, state licensure, staffing, and full operational launch
🚀 HOPE regulatory launch — only vendor ready at the CMS October 2025 deadline, serving 1,600+ post-acute providers across 48 states
AI-native product practice — built working healthcare AI prototypes using Claude and AWS Bedrock, plus Automated Attrition Protection — a churn prediction model built across enterprise post-acute platforms using CareCore EHR data ($7.7M revenue at risk identified, 94% accuracy, executive buy-in)
🔗 Platform architect — 35+ application integrations, 40+ vendor APIs, centralized validation engine serving 700+ CMS rules across care settings
Work artifacts
Career timeline
2022
– 2026
Principal Product Manager
Simple (Netsmart) · Home Health, Hospice & SNF Analytics
Identified HOPE as a market-defining regulatory inflection point and pivoted Simple's platform to hospice — launching the only HOPE-ready solution at the October 2025 CMS deadline, serving 1,600+ post-acute providers. Centralized CMS scrubber rule ownership previously duplicated across 5 separate software teams, delivering validated logic to 4 EHRs. Built unified Home Health / Hospice API platform, integrated directly into myUnity Essentials and Enterprise.
2014
– 2022
Sr. PM → Principal PM
Strategic Healthcare Programs (SHP) · CAHPS & Quality Analytics
Managed the largest HHCAHPS and CAHPS Hospice programs in the country — ~60% HHCAHPS market share, ~40% CAHPS Hospice. Piloted mixed-mode data collection that influenced CMS methodology. Led the first NHIA/NHIF Medicare Part B pilot in the country. First platform to integrate quality data with patient satisfaction benchmarking across the post-acute continuum.
2012
– 2014
Senior Product Manager
MatrixCare · Assisted Living Clinical & Financial Platform
Recruited directly out of Columbine to replicate enterprise HIT leadership at scale. Owned end-to-end product strategy for Assisted Living clinical and financial modules. Led new AL clinical product from concept through market delivery, including make-vs-buy analysis. Served as acting Release Lead coordinating product, engineering, QA, and training across quarterly releases.
2004
– 2012
Director of Health Information Technology
Columbine Health Systems · 15+ Care Sites & Service Lines
Why this matters
Most PMs learn healthcare software from the vendor side. At Columbine, I owned all health information technology across SNF, ALF, IL, Home Health, Home Infusion, DME, long-term pharmacy, and outpatient therapy — every care setting, every system, clinical and financial. I didn't just understand these workflows academically. I sat in those buildings, trained the staff, gathered feedback from nurses at midnight, and configured software to match how care actually happened. That's the foundation everything since has been built on, and it's what makes my AI product thinking grounded in operational reality rather than theory.
Managed 35+ clinical, financial, and operational applications across the full post-acute continuum. Led enterprise MatrixCare EHR implementation across all facilities, Return-to-Acute monitoring design, and admission workflow redesign. Served as Privacy Officer for 1,600+ employees, overseeing all electronic PHI security policies. Delivered 700+ employee training clinics for system adoption across clinical and operational roles. Also opened two skilled nursing facilities from scratch — site selection, state licensure, staffing, and operational launch. Recruited by MatrixCare upon departure to replicate this work for their client base.
Open to conversation

If these prototypes raise questions,
I'd like to answer them.

Exploring product leadership roles where deep healthcare domain expertise, AI-native thinking, and fast delivery all matter. 20+ years in healthcare technology across both sides of the table — 8 years as a licensed operator, 12+ years as a product leader — and a genuine belief that AI in care delivery is one of the highest-leverage places to build right now.

[email protected]  ·  970.231.2234  ·  Available for a conversation this week