"Give me a lever long enough and I shall move the world." - Archimedes
An automated analyst you can trust.
We build the semantic layer and the guardrails behind it: agreed definitions for your business, connected data, and checks on the answers. Analytics and AI for healthcare operators, from someone who has run the numbers at national scale.
The problem
Your team is the integration layer.
Growing companies run on people acting as human middleware: copying between systems, rebuilding the same report, arguing about whose number is right. Then AI arrives and makes it worse, because the machines inherit the mess.
Garbage in
Five versions of every number
CRM, finance, contracts, and operations each keep their own truth. "Revenue" means three different things depending on who pulled it, and the board deck reconciles none of them.
Garbage in
Definitions live in people's heads
The one analyst who knows why Q2 was restated is the single point of failure for every answer. When they're out, the company is out.
Garbage out
AI that confidently summarizes the mess
Point an agent, a copilot, or a chat-with-your-data tool at ungoverned tables and it doesn't fix the garbage. It launders it into fluent, plausible, wrong answers.
Garbage out
Decisions wait on assembly
Every leadership question kicks off a week of exports and spreadsheet surgery. By the time the answer arrives, the decision already got made on gut.
The model was never your bottleneck. The substrate is. Fix the in, and the out takes care of itself.
How it works
Answers start in week one. What changes over time is who produces them.
We start as your embedded analyst: your questions answered, definitions written as we go. Underneath, we build the automated analyst that takes over the recurring work. Reports move over one at a time, as each earns trust. Then we go lighter.
Us, your embedded analyst
Questions answered. Definitions written as we go.
The automated analyst
Morning report, alerts, follow-up questions
Blueprint
We answer your questions.
Build
Reports move over as each earns trust.
Operate
It runs recurring work. We go lighter.
What makes the analyst trustworthy
Semantic layer
One place that says what every number means.
What each number counts, how the data connects, and what the number is for. People and tools work from the same agreed definitions.
Guardrails
Answers your team can check.
The analyst shows its sources and flags data gaps. We maintain quality checks and access controls, review the answers, and update the definitions as the business grows.
Yours from day one
No dependency on us.
Definitions, documentation, and the system belong to you. Every blueprint is written so another team can take it forward. Ongoing support stays your choice.
Scale the business without scaling the back office.
Decision latency: a week, then the same morning
Leadership questions answered by the standing report and guarded self-serve, not by a fresh spreadsheet project.
One version of every number
Board deck, dashboards, and AI answers all computed from the same governed definitions. Arguments about whose number is right simply end.
Analyst hours moved from assembly to judgment
The copy-paste layer disappears. The people you have spend their time on the decisions only people can make.
Headcount decoupled from growth
The analyst absorbs the reporting, reconciliation, and follow-up work that would otherwise be your next back-office hires.
These are the design targets every engagement is scoped against. Specific, measurable versions of them go in your blueprint, and the build is held to them.
Engagements
Start with the numbers you steer by.
The Blueprint turns those priorities into agreed definitions, a design for the analyst, and a practical build plan. You get useful analysis while we develop it. Fixed scope first, retainer second, dependency never.
Phase 1
Blueprint
A four-to-six-week discovery and design sprint. Embedded analysis on your real numbers, shared definitions, and the implementation plan.
Fixed fee. Standalone deliverable.
Phase 2
Build
We implement the roadmap: data foundation first, systems of record wired in, then the automated analyst on top. Working software every two weeks.
Monthly retainer. Less than one senior hire.
Phase 3
Operate
We maintain and extend the analyst as you scale: new sources, new automations, new questions. Or your team runs it and we step back.
Reduced retainer. Month to month.
What the Blueprint delivers
- A map of your data. How your CRM, financials, contracts, and operational KPIs connect, and where the gaps are.
- The first agreed definitions. The priority numbers, how each is calculated, and what each includes and excludes.
- A design for the analyst. The initial reports and questions, data connections, tool choices, access controls, and quality checks.
- A sequenced roadmap with estimates. What to build first and what it takes to run and maintain it.
The work is yours from day one.
Definitions, documentation, and the system. The Blueprint is written so another team can take it forward. Ongoing support stays your choice.
What we need from you
Two hours in week one to agree on the priority numbers. About an hour with each member of the corporate team to understand the work. Read access to the existing systems. After that we come to you with answers, not questions.
Patient data
Healthcare work usually means patient-level data and PHI. Before we access any of it: a business associate agreement, appropriate infrastructure and access controls, and any required vendor BAAs. Clinical platforms and EMR interfaces are scoped separately.
Founder
Built by an operator, not a slide deck.
Spencer Kline
Founder
Data science, AI rollouts, healthcare analytics
Erie, Colorado
- 15 yrs
- Analytics foundation first, AI on top
- 2 to 50+
- Markets scaled at DispatchHealth
- $25M to $1.7B
- Valuation during his tenure
- 4 to 39
- OI Infusion sites on the backbone he built
Fifteen years turning clinical, operational, and marketing data into strategy and the systems behind it. He builds the stack himself.
SVP of Analytics and Data at OI Infusion
2023 to present$250M business, 4 to 39 sites. Built the warehouse, the BI layer, and the AI agents behind it.
VP, Growth & Operations at DispatchHealth
2017 to 20232 markets to 50+, $25M to $1.7B. Created the supply/demand-matching algorithm.
Founder-operator at Your Home Senior Living
25-bed assisted living, run on an AI operating system he built.
VP, Life & Home Insurance at EverQuote / Cogo Labs
2011 to 2017Analyst to GM through the pre-IPO run.
Bring the short list of numbers you steer by.
30 minutes. Bring the numbers you run the business on, or the report nobody trusts. We'll show you what a grounded, automated analyst looks like on your data, and what it takes to build. We start there.
Or email admin@onelever.ai