Who we serve

Proof, one engagement
at a time.

Our case studies for health plans (regionals, MA/SNP), PE sponsors and their portfolio companies, state and national governments, and consumer products.

Healthcare · PE portfolio company · +$75M ARR · payer contracting

PE-backed multi-state behavioral health provider

In production 3 weeks Locally-hosted models
$7M+
in contract revenue identified
avoided additional FTE hiring

The operator

The problem we took

Three merged entities, hundreds of payer contracts, and no formal contract system or negotiation engine. A two-person team ran the payer lifecycle on PDFs and spreadsheets, with revenue leaking through missed negotiation windows, passed renewals, and reimbursement schedules the team could not reconcile against each contract.

Clients' people

What was already in the client's environment

The payer book they already had: 1,000+ agreements across drives, boxes, and laptops. Revenue cycle, operations, and finance set the operational priorities, with SharePoint as the primary repository. The client keeps the repository, the renewal matrix, and the final say at every checkpoint.

The build team

What shipped, what it returned

Live in three weeks: one repository on a 42-field schema, three entities reconciled, six agent workflows. Negotiation cases draft themselves 120 days before renewal.

$7M
opportunity identified
40+ → 8 hrs
contract ops per month
6 agents
rate to negotiation history
Healthcare · PE portfolio company · +$125M ARR · supply chain & inventory visibility

National supply chain, logistics, and security data company

In production 2 weeks to deployment Frontier models
~$0.5M
in planned PMO cost avoided over the engagement
full-team status 8 hrs/wk to 2, agent-led async

The operator

The problem we took

A new product launching with two major medical device manufacturers: four organizations, 15+ teams, $100M projected at stake for the primary buyer. A traditional PMO would add headcount and still miss the cross-functional risk.

Clients' people

What was already in the client's environment

Six workflows stood up on the internal tools the team already ran: Asana, Jira, Claude, Box, and the Microsoft suite. The PMO lead owns exceptions and the risk register. Every human correction teaches the system.

The build team

What shipped, what it returned

An AI-native strategy management office, live in two weeks: three workflows at the market, three at the company, a human checkpoint on each. Board and customer materials generate from the live operating state.

$150-200K
planned PMO hire avoided, per year
8 → 2 hrs
full-team status, agent-led async
3+ yrs
engagement run-rate
Healthcare · PE portfolio company · +$225M ARR · digital health marketplace

National digital health marketplace

In production 12 weeks to deployment Locally-hosted + frontier models
$3-7M
projected annual revenue from matching accuracy
90% of current bookings affected by this initiative

The operator

The problem we took

Up to 40% of users picked the wrong insurance plan and 7% could not find theirs, driving cancellations, provider churn, and missed revenue. Underneath, a decade-old database of ~12,000 plan records patched by manual research, offshore teams, and outside vendors.

Clients' people

What was already in the client's environment

The data-operations team moved from manual maintenance to high-judgment review. Product, operations, engineering, and data took the handoff with taxonomy, SOPs, and training, on the systems they already ran: their internal proprietary database and data lake, and the Microsoft suite. The growth came from data the business already owned.

The build team

What shipped, what it returned

A five-tier payer hierarchy rebuilt under the live marketplace with zero record loss, wired to the Aequalis Payer Insights Platform and a national claims clearinghouse.

$3-7M
projected annual revenue
~4,000
plans added, 2,000+ removed
90%+
of bookings covered
Public sector · state health agency · $7.5M annual license

Statewide readiness platform, via a national logistics data company

In production Ongoing fractional AI leadership Locally-hosted models
$15M
state contract secured for statewide deployment

The operator

The problem we took

A proven logistics, inventory, and supply-chain platform, redesigned and deployed for healthcare: statewide emergency preparedness and readiness across 8 regional coalitions, 8 health systems, and 2 state departments. Federal expansion is the long-term goal.

Clients' people

What was already in the client's environment

The client's existing platform did the heavy lifting, adapted rather than rebuilt. State health and technology departments, coalitions, and health systems align on one operating picture. We stayed on as the fractional AI arm, applying AI principles and practices to the system they already run.

The build team

What shipped, what it returned

The commercial case, ROI model, and pricing for a $15M state contract, then an AI-native deployment across every coalition and both departments.

$15M
state contract secured
5,000+
SKUs visible statewide
8 · 8 · 2
coalitions, systems, state depts
Healthcare · PE portfolio company · +$250M ARR · clinical data intelligence

Clinical data intelligence company entering risk adjustment

Board delivered 8 weeks to delivery
$60M+
projected annual uplift for a new line of business

The operator

The problem we took

Provider-market dominance, and a board mandate to take a risk adjustment product into payer segments the commercial team had never sold to. Started from no payer value proposition and no target list, in a crowded and maturing market.

Clients' people

What was already in the client's environment

Built around the client's own product, growth, and sales teams, with 15 voice-of-customer interviews reached through the 3Pillars network, VP to board level. The sales team kept and shaped the growth enablement kit; the board approved the roadmap.

The build team

What shipped, what it returned

County-level MA enrollment and risk-score datasets sized the opportunity target by target. Buyer-specific messaging, PMPM ROI models, and a board-level recommendation in 8 weeks.

15
VoC interviews, VP to board
10+
competitors profiled
7
deliverables, 2 board updates
Public sector · state government

Statewide police force, every district

In production Ongoing fractional AI leadership Locally-hosted models
80% less
senior-officer time per case review, measured in production
30-45 min down to 5-10 · 10,000+ hours returned

The operator

The problem we took

Senior officers spent 30 to 45 minutes on every case review, and situational awareness was assembled by hand, district by district, hours late. Modernization had to happen inside the government's own environment.

Clients' people

What was already in the client's environment

The state's existing infrastructure and camera estate, kept: video intelligence runs above installed CCTV without replacing it. Officers' corrections retrain the models, and command reviews every flag.

The build team

What shipped, what it returned

Seven interconnected production systems on locally-hosted models: grievance processing, digital evidence, investigative data fusion, document digitization, video intelligence.

5-10 min
case review, was 30-45
10,000+
man-hours returned
40%
faster incident response
Our insights

The thinking behind the work, by industry.

All insights →
Health plans

Three levers for AI in a health plan

Most plans point AI at the smallest lever. The bigger returns sit in the other two.

4 min read
Health plans

The office that actually pulls the levers

Naming the levers is easy. Pulling them is a question of structure.

4 min read
Cross-industry

Consolidate first, then deploy

A hundred workflow variants collapse into a handful, in the right order.

4 min read
Health plans

Health plans already ran this experiment

VBC taught plans what happens when silos own the roadmap.

3 min read
Investors

The $600K math

A credible AI hire costs $600K. The roadmap problem survives the hire.

3 min read
Investors

You need a ledger, not a metric

Activity dashboards don't answer the board. A benefits ledger does.

3 min read
Health plans

The speed trap

Two versions of prior auth: one runs faster, one changes the work.

4 min read
Investors

The moat inversion

When every buyer has the same models, your context becomes the moat.

3 min read
Cross-industry

Own the layer that stays

Route workloads across models. The context layer, locally hosted, is yours.

3 min read
Health plans

The layer tax

Payer data arrives in layers. Each one caps what AI can return.

3 min read
Health plans

Every AI roadmap is secretly a data roadmap

The AI roadmap slips when the data roadmap under it was never written.

3 min read
Health plans

The talent auction

Regional plans cannot win the $600K talent auction. They do not need to.

3 min read
Investors

The office agents report to

Agents that cross silos need an office that owns them.

3 min read
Health plans

The forward-deployed health plan

Sort the org chart into volume and judgment, then staff to the second.

3 min read
How we AI

Most regional MA exits are not exits

Most "exits" are sales. Contract-level tracing tells them apart.

4 min read
How we AI

Reading the Cityblock-Homeward deal

The read the headlines missed, built from public CMS data in an afternoon.

4 min read