Outcomes that were built, not projected.
I led every engagement directly with the client team, from diagnosis through build to measured outcome. Client identities are anonymized; every figure below is real.
- Recovered, reallocated, or renegotiated
- $18M+
- Documented engagements
- 5
- Measured outcomes
- 20
- Healthcare verticals
- 5
Engagement details anonymized to protect client confidentiality.
64%
After-hours inquiry-to-admission rate
An AI voice system answered crisis calls when no human could, and flipped inquiry-to-admission from 18% to 64%
After-hours inquiry-to-admission rate
0%
Swipe to explore →
The problem
A 14-facility behavioral health network across the Southeast was hemorrhaging after-hours inquiries. Their crisis line rang unanswered on nights and weekends: the exact moments when families, desperate and ready to commit, finally made the call. Admissions coordinators would return voicemails the next morning, but by then half the callers had either talked themselves out of treatment or found another provider. Internal data showed 62% of after-hours calls never converted to a scheduled assessment. Marketing was spending $340,000 a month driving inbound volume that operations could not capture.
The build
I spent three days listening to recorded calls, both answered and missed, to map the emotional arc of a family in crisis reaching out for help. I then designed a conversational AI voice agent trained on the clinical intake criteria their own admissions directors used, deployed it as the first responder for all after-hours and overflow calls, and connected it directly to the CRM so every interaction created a qualified lead record with a sentiment score, urgency flag, and recommended follow-up window. The system escalated high-acuity calls to an on-call clinician within 90 seconds while handling routine inquiries conversationally and booking assessments directly on the calendar.
The result
Within the first full quarter, after-hours inquiry-to-admission conversion jumped from 18% to 64%. The AI system captured 1,847 calls that previously would have gone to voicemail, generating 1,182 scheduled assessments and 684 admissions: $2.7M in incremental revenue at the network's average reimbursement rate. The admissions team, freed from playing voicemail catch-up, reduced first-contact-to-assessment time from 38 hours to under 4 hours across all channels.
The readout
$2.7M
ResultIncremental revenue from recaptured after-hours calls
0
ResultPreviously-missed crisis calls captured per quarter
0h
DownFirst-contact-to-assessment time, down from 38h
He told us on day one that AI would not fix a broken workflow. He was right. He made us fix the workflow first, then applied automation where it would actually create leverage. Our intake response times improved materially.
89%
New patients traced to source channel
A 97-location dental group was spending $8.2M on marketing with no idea which patients came from where. I fixed it in six weeks.
The problem
A 97-location dental group operating across Texas, Colorado, and Arizona ran campaigns across Google Ads, Meta, TikTok, direct mail, radio, and community events: $8.2M in annual spend distributed across 14 agencies and platforms. But their patient management system had no attribution fields. New patients simply appeared in the schedule. When the CMO asked each agency to prove ROI, she got fourteen different versions of the truth, none of which matched the actual production reports from the practices. The CFO was ready to cut the marketing budget by 40% at the next board meeting.
The build
I built a lightweight attribution layer that sat between their marketing channels and their practice management system without requiring a full PMS migration. Every ad platform, landing page, phone number, and intake form was instrumented with source tracking that persisted through the entire patient journey, from first click through new patient exam to treatment plan acceptance. I configured their CRM to surface source-channel contribution at the patient, practice, and region level, and established a weekly attribution review that the CMO, CFO, and regional operators attended together with a single dashboard as the source of truth.
The result
Within six weeks, the group could trace 89% of new patients to their originating channel and campaign. The data revealed that TikTok and community events were driving 3.2x more new patient starts per dollar than their largest agency retainer. They reallocated $1.7M from underperforming channels into higher-yield programs, and new patient volume increased 31% year-over-year without increasing total marketing spend. The CFO withdrew the budget-cut proposal and approved an additional $600K for the channels the data validated.
Attribution coverage
From near-zero to 89% of patients traced to source.
$8.2M of spend once vanished into fourteen versions of the truth. Once every patient carried its channel, the money followed the yield.
$0.0M
Reallocated from low-yield channels to high-yield
+0%
New patient volume, same total spend
The readout
$1.7M
ResultBudget reallocated from low-yield to high-yield channels
0%
GainIncrease in new patient volume, same total spend
0
ResultWeeks from kickoff to full attribution visibility
We brought Kevin in to fix attribution. He ended up redesigning how marketing, admissions, and clinical teams hand off responsibility. The measurement improvement was a byproduct of fixing the operating model.
$4.3M
Annual revenue leakage identified and recovered
Connecting the EHR, the CRM, and the marketing stack for the first time, and finding $4.3M in invisible leakage
The problem
A three-hospital network with 22 outpatient clinics in the Pacific Northwest operated with a technology stack that had grown organically over fifteen years. Their Epic EHR, Salesforce CRM, Marketo instance, and call center platform had never been integrated. Patient referrals from primary care physicians were faxed. Marketing campaigns generated leads that were emailed to clinic managers as PDFs. Service-line profitability was calculated once a year by a team of financial analysts who spent six weeks reconciling spreadsheets. No one in the system could answer a simple question: when a patient scheduled a cardiology appointment, what sequence of events (referral, campaign, web search, physician recommendation) actually caused that appointment to happen?
The build
I ran a 360-degree revenue operations diagnostic across their entire technology and process stack, mapping every system, data handoff, and workflow that touched a patient from awareness through appointment to billing. I identified 31 disconnection points where data was either lost, duplicated, or manually transcribed. I then architected an integration layer connecting Epic, Salesforce, Marketo, and the call center platform through a centralized patient identity resolution engine, built a service-line revenue dashboard that updated daily instead of annually, and trained each department on the new operating cadence.
The result
The integration surfaced $4.3M in annual revenue leakage: referrals that were never scheduled, prior authorizations that expired before the appointment, and services delivered but never billed. Service-line profitability reporting went from a six-week annual exercise to a daily dashboard. The cardiology service line alone identified a 22% referral leakage rate from their largest PCP network and recovered $1.1M in missed appointments within the first quarter after fixing the handoff process.
System integration
Four systems that had never spoken, wired into one.
Fifteen years of tools bolted on side by side. Connecting them through a single patient identity resolved what each system was quietly dropping on the floor.
Swipe to explore →
$0.0M
Annual leakage surfaced by the integration
0
Disconnection points across four systems
The readout
0
ResultDisconnection points identified across four core systems
Daily
ResultService-line profitability reporting cadence, from every 6 weeks
0%
DownReferral leakage rate corrected in cardiology alone
We had more tools than we could count and still could not answer basic questions about where patients were coming from. Kevin untangled the stack, connected what mattered, and gave us a dashboard leadership actually trusts.
$7M
Purchase price reduction from operational findings
A pre-acquisition operational audit that found $2.3M in hidden revenue leakage before the deal closed
The problem
A healthcare-focused private equity sponsor was 45 days from closing on a $140M acquisition of a multi-site autism and specialty care platform with 28 clinics across four states. The financial diligence was clean: revenue was growing 18% year-over-year, margins were in line with industry benchmarks, and the add-on pipeline looked strong. But the sponsor's operating partner had a gut feeling. The platform's revenue cycle metrics (days in A/R, denial rates, collection rates) were reported at the corporate level with no clinic-level visibility. Provider utilization data was maintained in a spreadsheet that the clinical director updated monthly by emailing each clinic manager. The investment committee needed to know: were these numbers real, and could the operating model actually scale to the 60-clinic target in the investment thesis?
The build
I embedded with the target company's operations team for three weeks. I audited their revenue cycle end-to-end (insurance verification, authorization, claims submission, denial management, and patient collections) across all 28 clinics at the transaction level. I shadowed providers to measure actual clinical hours against scheduled hours, mapped their intake workflow from first call to first appointment, and stress-tested their technology stack against the volume projections in the acquisition model.
The result
The audit uncovered $2.3M in annual recurring revenue leakage: $1.4M in denied claims that were never reworked, $620K in services delivered but never coded due to a documentation gap between providers and billers, and $280K in patient balances older than 180 days with no collection process. I also identified that provider utilization, reported at 87%, was actually 71% when measured by billable clinical hours, a gap that would have made the 60-clinic roll-up thesis unachievable without operational restructuring. The sponsor used the findings to negotiate a $7M purchase price reduction and funded a six-month operational turnaround plan that closed the leakage before the first add-on acquisition.
Diligence findings ledger
Clean on paper. $2.3M of leakage underneath.
Corporate-level reporting hid the detail. Auditing the revenue cycle at the transaction level, across all 28 clinics, surfaced what the financials never showed.
Swipe to explore →
$0M
Purchase price reduction the findings justified
Verified provider utilization vs. what was reported
The readout
$2.3M
ResultHidden annual revenue leakage uncovered pre-close
0%
DownActual provider utilization vs. 87% reported
0
ResultClinics audited at transaction level in three weeks
The difference is he has actually run a company. He understands P&L pressure, board expectations, and what it takes to get a team aligned behind a plan. The strategy advice comes from having done it.
90 days
From 64% to 90%+ occupancy
Three underperforming senior living communities, a 90-day sprint, and a marketing and intake engine that filled every vacant unit
The problem
An 11-community senior living operator had three Midwestern communities running at 64% occupancy, well below the 89% breakeven threshold. Each community had its own marketing approach: one was running Facebook ads managed by the activities director, another relied entirely on hospital discharge referrals, and the third had no active marketing at all beyond a website that had not been updated since 2019. The corporate marketing director was spread across all 11 communities and had no budget for agency support. The board had given the CEO 120 days to show a turnaround plan or begin the process of selling the three communities at a loss.
The build
I built a centralized intake and marketing engine that served all three communities from a single operations hub. I stood up localized paid search and social campaigns targeting adult children of seniors within a 25-mile radius of each community, designed landing pages that converted at 4x the old website, deployed an AI phone agent to handle after-hours and weekend inquiries with immediate tour scheduling, and configured a lightweight CRM that gave each community director a real-time view of their pipeline (inquiries, tours, deposits, move-ins) without requiring any technical expertise to operate.
The result
Within 90 days, all three communities reached 90%+ occupancy for the first time in three years. The centralized engine generated 412 qualified inquiries, scheduled 287 tours, and converted 94 move-ins across the three communities. Cost per move-in averaged $1,840: less than half the industry benchmark of $4,200. The CEO presented the model to the board as the blueprint for all 11 communities, and the operator allocated budget to roll out the centralized intake engine across the full portfolio within six months.
Aggregate occupancy across 3 communities
0%
Swipe to explore →
The readout
0
ResultDays to fill three underperforming communities
0
ResultMove-ins generated from 412 qualified inquiries
$0
DownCost per move-in vs. $4,200 industry benchmark
0
NewCommunities in the blueprint adopted for full rollout
Kevin does not operate like a typical consultant. He builds the system alongside your team, trains them to run it, and holds himself accountable for whether it actually works. That is rare.
See how this applies to your operating constraint.
Bring the constraint limiting your next stage of performance. I will help you determine the most practical next move, and stay accountable through the build.