Preventive-Health Screening for Self-Insured Employers
Most wellness programs can’t prove they work.So we built one that doesn’t pretend to.
Signal Labs is a 60-minute on-site preventive-health assessment that turns strength, movement, body composition, aerobic capacity, and cognition into one deterministic 0–100 Signal Score — with a Red / Amber / Green readout and an early-decline alarm that flags a downward trend before symptoms appear. Built for self-insured employers who are done being sold fake ROI.
No individual data ever reaches HR. Every claim on this page is sourced or labeled a projection.
The exposure nobody names at the renewal meeting
You find decline too late — and you’re holding the data either way.
Problem
You’re flying your self-insured plan on lagging indicators.
By the time a high-cost claim shows up in your trend report, the decline that caused it has been building for months — sometimes years. Your current stack measures steps and sign-ups, not whether your workforce is actually getting stronger or weaker.
Agitate
The wellness data you DO collect can become a liability.
Individual health information sitting in a vendor’s database is HIPAA, ADA, and GINA exposure with your name on the contract. So you’re exposed twice: you find decline too late, and you’re holding sensitive data you’d rather not be responsible for. Meanwhile the last vendor’s “$3.27 saved per $1” promise didn’t survive the first hard question from your CFO.
Solution
There’s a version of this where you see the trend early — and never touch an individual’s health data at all.
You see the trend early, you can defend the spend, and HR never holds an individual’s health record. That’s what we built.
From lagging to leading
The number is math you can audit. The explanation is the friendly part.
Lagging claims data, point-solution shelfware, and a privacy footprint you’d rather shrink.
A clear, deterministic Signal Score per consenting employee — Red / Amber / Green — with a statistically rigorous early-decline alarm. Employees own their results. HR sees only privacy-protected aggregates, never an individual.
Signal Labs fuses six measured domains into a transparent 0–100 score using deterministic math — not a black-box guess. An early-decline alarm uses Kalman filtering and CUSUM change-detection to flag a downward trend before it becomes a symptom. A large language model only narrates the result in plain language — it never calculates the score.
Low-friction by design
Four steps. One cohort. No org-wide bet.
01
Book a pilot cohort.
A limited group of consenting employees, on-site, ~60 minutes each. No org-wide commitment required.
02
Measure six domains.
Trained staff capture strength, power, body composition, movement, aerobic capacity, and cognition.
03
Compute the Signal Score.
Deterministic 0–100 (Physical + Cognitive), Red / Amber / Green, with a Kalman/CUSUM early-decline alarm. The LLM narrates; it never scores.
04
Deliver results — to the right people.
Each employee privately receives their own result. HR receives only k-anonymous, differentially-private aggregates. No individual data crosses the line. Ever.
Concrete, not vague
Six measured domains, fused into one auditable score.
Strength
Functional force production; a foundational marker of healthspan.
Power
Rate of force; among the earliest capacities to decline with age.
Body composition
Lean mass vs. fat mass, beyond the blunt instrument of BMI.
Movement
Mobility and quality of movement patterns.
Aerobic capacity
Cardiorespiratory fitness, a core longevity signal.
Cognition
Processing and function, narrated in plain language.
Each domain feeds the deterministic Signal Score. Optional research-grounded biological-age context is available using published methods — Klemera-Doubal (2006) [V] and DunedinPACE (2022) [V]. General-wellness only; non-diagnostic.
The screen that catches what you'd otherwise miss.
Most health apps show you yesterday. Signal Labs shows you the direction you're heading — a deterministic 0–100 score, an early-decline alarm, and the published math behind every number. Everything on this page is the actual shipped app.
- Deterministic 0–100 score
- Early-decline alarm
- HR never sees an individual
If you measure it, you can move it. Individuals get the whole picture on day one. Teams get privacy-protected aggregates — and not a single individual row.
The product, not a pitch
How Signal looks in the app and on the web.
One number you can actually trust
Your whole stack — wearables, labs, body composition, cognition — fused into a single deterministic 0–100 Signal Score with a Red / Amber / Green readout.
- Computed by published, versioned math — not a black box
- Red / Amber / Green so you know when to act
- The same score on web and on your phone
The one signal no tracker measures
A 6-minute, clinically-validated cognitive battery folded into your score — the brain-health measure every wearable skips.
- Reaction time, attention, processing speed & more
- Tracks change over time, surfaces drift early
- Matters most if decline runs in your family
See decline before symptoms
A Kalman filter with CUSUM change-detection watches your trajectory and flags a downward trend while you can still act on it.
- 30-, 90- and 365-day views of every signal
- Early-decline watch on the metrics that matter
- Estimates carry a stated confidence band — never false precision
A companion that remembers — and never decides
Ask it anything about your health in plain language. It reads your own numbers, explains what they mean, and remembers what matters to you between conversations — so the advice gets more useful over time, not more generic.
Remembers what matters
Your goals, your routine, and the interventions that actually worked for you — carried from one conversation to the next, so you never start from scratch.
You stay in control
View, edit, or erase anything it remembers, any time. Your memory is yours — it’s there to help, not to lock you in.
Explains, never decides
It puts your deterministic results into plain language. It cannot move your score, change a traffic light, or raise an alarm — those are computed, not generated.
The science under the score
Authority is your currency. We hand you the evidence to win the budget fight.
Deterministic, not a black box
The same inputs always produce the same number. No model drift. The LLM only translates results into plain language — it has zero influence on the score.
Established statistical methods
The early-decline alarm uses Kalman filtering and CUSUM change-point detection to separate a real downward trend from normal noise.
Peer-reviewed bio-age context
Optional biological-age context uses published methods: Klemera-Doubal (2006) [V] and DunedinPACE (2022) [V].
The market context (sourced)
0%
of large US firms already offer a wellness program
$0B
corporate-wellness market — most of it can’t prove its outcomes
0%
of prior savings estimates ruled out by a landmark workplace-wellness RCT
What customers say
[pilot testimonial — to collect]
Benefits Director — [pilot employer — to collect]
[pilot testimonial — to collect]
VP Total Rewards — [pilot employer — to collect]
We’re early, and we won’t fake proof. Our pilot cohorts are how the evidence gets built — with your data governance, not around it.
Let’s kill the number you’ve heard a hundred times
We sell Value-on-Investment — and we show our math.
The “$3.27 saved per $1 spent” wellness ROI claim is discredited. The Illinois workplace-wellness RCT (Jones, Molitor & Reif, 2019) ruled out 84% of prior savings estimates [V]. If a vendor is still quoting that multiplier, they’re quoting a ghost.
VOI counts the things you can defend in front of your CFO and legal team:
- Early visibility into physical and cognitive decline trends — at the aggregate level.
- Reduced data-liability footprint (HR never holds individual health data).
- A program employees trust enough to use — because their results stay theirs.
- Workforce-performance and longevity signals leadership actually cares about.
Any forward-looking dollar figure we show you is labeled [P: projection] with its assumptions exposed. We would rather lose the deal than hand you a number you can’t defend.
ROI calculator — honest teaser
2,000 employees
$250,000
Pilot figure is [P: projection] — illustrative, capped at a 200-person cohort, assumptions visible. No inflated multipliers; the full calculator labels every output [V: source] or [P: projection].
Run the full ROI calculatorPrivacy by architecture
HR never sees an individual. This isn’t a policy. It’s the architecture.
k-anonymity
Aggregates are only ever reported in groups large enough that no individual can be re-identified.
Differential privacy
Statistical noise is added so the math can’t be reverse-engineered back to a person.
Row-Level Security (RLS)
Enforced at the database layer — “HR can’t see individual data” is a rule the system cannot break, not a promise a person keeps.
BAA in place
We operate under a Business Associate Agreement; data handling is built to be HIPAA-conscious and general-wellness-scoped.
Employees own their results
Each person privately receives their own Signal Score. Their manager doesn’t. HR doesn’t. You don’t.
This is the part that makes the rest work: employees participate honestly because they know you’ll never see their number — and honest participation is exactly what makes your aggregate trustworthy.
The hard questions, answered
Objections we’d rather you raise now.
Stop measuring your workforce’s health too late.
You don’t need to bet the whole benefits budget to find out if this is real. Book one pilot cohort. Measure six domains, see the deterministic Signal Score and the early-decline alarm on your own people, and confirm — with your own eyes — that HR never sees an individual.
Pilot capacity is genuinely limited each quarter by our on-site assessment throughput. When a quarter’s cohorts are full, the next opening is the following quarter — an operational fact, not a countdown timer. Real scarcity only
You already vet vendors more carefully than most. Apply that same rigor here — and let the data, not our adjectives, make the case.