Cognitive health integration for occupational health software
Turn every occupational health visit into a decision teams can act on. Padoa Cognitive Health, powered by CogniFit, is a proposed module in three layers: a 3-minute Cognitive Resources Index for every employee, a full Cognitive Check when it matters, and personalized training and team actions, built on CogniFit's APIs and SDKs instead of a new engine.
3 minutes for the employee · 27 languages · normed on 3,214,998 assessments · SOC 2 and GDPR
Example data. Visible only to the employee and their occupational health team.
The case in 60 seconds
Padoa Cognitive Health is a proposed Padoa module, powered by CogniFit, that measures each employee's cognitive resources in three minutes, offers a full assessment when it matters and supports people with personalized training, without Padoa building its own cognitive engine.
A new revenue line on the installed base
Padoa resells the index to employers at a price it sets, and full assessments on demand. CogniFit runs the engine, the norms and the updates.
Data other platforms do not have
A longitudinal, norm-referenced measure of workforce cognitive resources, useful for Padoa's AI roadmap and its expansion beyond France.
More reach for scarce clinicians
An objective signal before the visit helps the occupational health team prioritise follow-up. The clinician decides.
Real data for the employer who pays
A quarterly team barometer for the psychosocial risk assessment, for groups of 20 people or more and only on an approved legal basis.
The request: a 12-month pilot with one occupational health service and 2,000 to 5,000 employees, with clear exit criteria.
About this page. A concept prepared by CogniFit: nothing is signed, built or launched, and Padoa already works on prevention and psychosocial health. People, scores and program figures are synthetic and marked Example data; revenue figures are Scenario assumptions.
Visits produce rich data about individuals, which rightly stays private. Employers plan for teams with little to go on.
Too few clinicians
Prevention duties keep growing while physician time does not. France recorded 74,602 mid-career visits in 2024.5
Blind to cognitive load
In France the occupational risk assessment must cover psychosocial risks, yet it rests on little objective data.
No useful feedback
They leave the visit with a certificate, not with an idea of what is costing them focus.
A workflow others can match
Scheduling and records are becoming standard. Data no one else has is what sets a platform apart.
Private-sector absenteeism in France in 2024, about €1,535 in direct costs per employee per year.6
Of GDP lost to insufficient sleep in the hardest-hit of five OECD countries studied, through absence and lower productivity.8
Of dementia cases linked to 14 modifiable risk factors, many of them everyday habits.9
How it works: three layers, one product
Everyone takes the short index. Only those who need it go further. Training and team actions follow, and the next quarterly index shows what changed.
The next quarterly CRI measures the change
Example data for one employer, not measured outcomes.2
- Layer 1 · everyone
Cognitive Resources Index (CRI)
Three minutes inside the pre-visit questionnaire employees already fill in, then every quarter from their personal space. It measures cognitive resources and habits: the employee gets their levers, the team gets an objective signal, the employer gets a team barometer.
- Layer 2 · when needed
Padoa Cognitive Check
If the CRI comes out Amber or Red, or the professional decides, the full CogniFit assessment (about 20 minutes) is requested from the medical record and followed between visits.
- Layer 3 · action
Padoa Brain Health and team actions
Personalized cognitive training for the employee, team actions for the employer, both measured by the next CRI.
Later, as a separate conversation: Padoa Cognitive Insights, de-identified program evaluation and research under independent governance.
Inside the three minutes
Padoa calls CogniFit's API with a pseudonymous ID, so CogniFit never knows who the person is.
1 · In Padoa, about 3 minutes
Seven quick questions on one screen, then a 90-second symbol–digit task.
Your week
2 · The CogniFit CRI engine: from raw answers to a calibrated index
Normative base of hundreds of thousands of people who completed these tasks, by age, sex, education, language and device.
- Device correction. Removes screen and touch latency, so phones and laptops are comparable.
- Norm referencing. Compares the result with people of the same age, sex, education and language.
- Resource weighting. Combines habits and performance with weights learned from the norms.
- Personal baseline. From the second check on, compares people with themselves and discounts practice effects.
- Quality control. Flags rushed or interrupted sessions instead of scoring them.
3 · Back into Padoa
- AEmployeeTheir index, their two or three levers and practical next steps.
- BOccupational health teamFull detail in the medical record, with a suggested follow-up.
- CEmployer and HRA team barometer only: trends, drivers and actions, for groups of 20 or more.
Resources in good shape
Habits and performance at or above the person's norm and baseline.
A lever needs attention
One weak habit, or performance slightly below the person's expected level.
Resources running low
Several weak levers and a reliable drop against the person's baseline. Offered as a conversation with the occupational health team.
The CRI is a prevention indicator. It does not diagnose, does not assess fitness for work and is never used to hire, promote, evaluate or discipline anyone.
An Amber or Red result opens the way to a full Cognitive Check, requested from the record. See the four steps in the clinician view.
Not another app. A new capability inside Padoa.
Explore how the three layers could fit the day of an occupational health clinician, an employee and an employer. Every person, score and figure below is example data.
Occupational health professional
Example dataCamille D., 44
Routine occupational health follow-up
Dr. Sophie Laurent
Cognitive health overview
- Apr 202671
- Jul 202668
- Oct 202665Amber
Suggested: a follow-up Cognitive Check. Sleep is the weakest lever and processing speed has dropped against Camille's own baseline for two quarters. This helps the team prioritise follow-up; the clinician decides.
Suggested follow-ups, each one left to professional judgment: discuss the profile with the employee, propose voluntary training, plan a repeat assessment at a later visit.
Request a cognitive assessment
- Step 1 of 4
Select the assessment or approved task
The real catalogue and its suitability for each workflow would be reviewed with Padoa's medical team.
- Step 2 of 4
Review purpose, employee and authorization
Purpose: follow-up of an Amber CRI and a cognitive concern raised in consultation.
Employee: Camille D., synthetic record.
Authorization: the legal basis and consent flow agreed for this workflow, confirmed before sending.
- Step 3 of 4
Confirm the secure invitation
The invitation would appear in Camille's Padoa employee space. No second account, no separate login.
- Step 4 of 4
Invitation ready
Demo simulation: no invitation was sent.
In the proposed product, the result would return to this record when Camille completes the assessment.
Cognitive profile · last full assessment
Staying on task and filtering out distractions.
Holding and using information for a short time.
How quickly simple information is taken in and handled.
Planning, switching between tasks and inhibiting responses.
Mock scores on a 0 to 100 scale. They are not clinical findings, and domains are not meant to be compared with each other.
Longitudinal results
Attention: 66 to 74 across four assessments, +8, with 18 training sessions in between.
Synthetic data for illustration. Other domains: working memory 63 to 68, processing speed 77 to 81, executive functions 70 to 76.
Training options
Illustrative plan for Camille: three short sessions a week, about 10 minutes each, weighted toward working memory and attention, reviewed at the next visit and measured by the next CRI.
CogniFit adapts training to each person's results. No particular outcome is promised.
Proposed experience using CogniFit APIs and SDKs. Exact workflow, assessment selection and record integration are subject to technical and clinical review.
Employee
Example data
Your individual cognitive results are private
- You see all your own results, and can export or delete them.
- Authorized occupational health professionals see what an approved care workflow requires.
- Your employer never receives your scores.
- Employer reports, where lawful, cover groups of 20 people or more only.
Your cognitive health, in your hands.
Discover your cognitive strengths, follow changes over time and use personalized exercises designed to challenge your cognitive skills.
Your Cognitive Resources Index this quarter
One lever needs attention: sleep. Your processing speed is a little below your own baseline.
Your next steps
- Keep a regular wake-up time for the next four weeks
- Try a short daily walk: physical activity is your second lever
- Talk to your occupational health team at your next visit, if you wish
Your index is a prevention indicator, for your information. It is not a diagnosis.
Discover your cognitive profile
Invitation from your occupational health service: a 20-minute online assessment, on your own schedule.
Attention is how well you stay on task. Working memory is how you hold information for a moment.
Your results are for information. They are not a diagnosis.
Your cognitive journey
Assessment completed · 18 sessions done · attention 66 to 74
Next: your quarterly index in January and a repeat assessment after your next visit.
Your next best activity
Today: 8 minutes of attention training
Illustrative suggestion, not a validated personal recommendation.
Find the arrow that points a different way. Round 1 of 5.
Interactive illustration, not a real clinical assessment or a CogniFit test.
Employer
Example data- Launch
- Invitations sent
- Opt-in window
- Ongoing program
Training sessions per month, all participants
| Site | Participants |
|---|---|
| Site A | 310 |
| Site B | 228 |
| Site C | Hidden: fewer than 20 people |
Not available in the employer view. Individual results stay with the employee and authorized occupational health professionals.
Prevention insights without exposing individual medical data. No names, no individual scores, no rankings. Every aggregate needs at least 20 people, an approved legal basis and validation by the occupational health service.
Want Padoa's product team to walk through this prototype with CogniFit?
Who sees what: people see themselves, clinicians see patients, organizations see teams
Individual results stay inside a protected boundary. Only aggregated, thresholded data crosses it.
Protected individual data
Aggregated data only, groups of 20 or more
| Recipient | Receives | Never receives |
|---|---|---|
| A · Employee | Their CRI, levers, trend, full profile and training | Anyone else's data |
| B · Occupational health team | Full individual detail, in the medical record, within an approved care workflow | Data outside that workflow |
| C · Employer, HR | Team shares, trends, drivers, suggested actions, participation and engagement | Any individual result, names linked to results, rankings, risk lists, or any group under 20 people |
| Padoa | Results to display and store under its health data arrangements, as proposed | Use beyond the approved purpose |
| CogniFit | Pseudonymous answers and task data | Names, contact details, employer identity |
Padoa with and without CogniFit: same platform, a different product
Occupational health already follows many aspects of worker health over time. Cognitive resources could become one more measured, longitudinal dimension, inside the workflow clinicians already use.
Without a cognitive layer
- A self-reported pre-visit questionnaireNo objective measure of cognitive resources.
- Cognitive concerns as free-text notesComparing one visit with the next depends on how each note was written.
- Every case competes for clinician timeLittle to help decide who needs follow-up first.
- Employers receive certificatesNothing about teams they can plan around.
- Workflow dataSimilar to what any competitor holds.
With Padoa Cognitive Health, as proposed
- An objective, norm-referenced measureIn three extra minutes, in 27 languages.
- A clearer view of who needs follow-upHelps the occupational health team prioritise follow-up; the clinician decides.
- Full assessments from the recordScores sit in the record and compare across visits.
- A quarterly team barometer for HRReal data for the psychosocial risk assessment, never individual results.
- A new add-on to resell, and a longitudinal datasetPriced by Padoa; a measure of workforce cognitive resources for its AI story and international growth.
Objective measures per employee per year: one at the visit, plus quarterly re-checks.
What HR can see, updated each quarter, where a legal basis allows.
What could Padoa Cognitive Health be worth?
Two views of the same scenario: what Padoa could earn on top of its platform, and what the employer who pays could get back. Scenario tools built on your inputs, not forecasts. All amounts in euros.
Revenue within the existing base
The index as an add-on for employers, full assessments on demand, and training programs, if willingness to pay is confirmed.
A broader prevention platform
New specialist capability for clinicians and a reason for employees to engage between visits.
Reusable in new markets
The same components could travel to other European markets, once localized for language, clinical rules and regulation.
Enter your figure. The starting value is a round placeholder, not a Padoa figure.
A hypothetical price set by Padoa, not an approved Padoa or CogniFit price.
CogniFit cost per full assessment
€4.60 per assessment up to 100 assessments a year, falling in a straight line to €1.15 from 100,000. A scenario assumption, not a CogniFit quote.
Source: scenario assumptions, note 2
Scenario: Padoa revenue per year
€750,000
Based entirely on the figures you enter. Not a revenue forecast or a CogniFit quote.
- Employees enrolled in the CRI100,000
- CRI revenue (enrolled × Padoa price)€600,000
- CogniFit CRI cost (enrolled × €3)€300,000
- Full assessments a year10,000
- Assessment revenue (assessments × Padoa price)€150,000
- CogniFit cost per assessment€4.26
- CogniFit assessment cost€42,600
- Margin before other costs€407,400
| Employees enrolled | Full assessments | CogniFit cost per assessment | Padoa revenue | CogniFit cost | Margin before other costs |
|---|---|---|---|---|---|
| 1,000 | 100 | €4.60 | €7,500 | €3,460 | €4,040 |
| 10,000 | 1,000 | €4.57 | €75,000 | €34,570 | €40,430 |
| 100,000 | 10,000 | €4.26 | €750,000 | €342,600 | €407,400 |
| 1,000,000 | 100,000 | €1.15 | €7,500,000 | €3,115,000 | €4,385,000 |
What this figure leaves out
The margin shown subtracts only the CogniFit costs used in this simulation. It is not EBITDA, contracted ARR or net value. Commercial due diligence would set final pricing, distribution rights, revenue share, implementation costs, customer acquisition and ongoing support. None of these are assumed here.
Scenario: estimated value per euro spent
4.2 ×
Productivity recovered is an estimated value, not a cash saving. Effects are assumptions the pilot would test.
- Productivity value recovered (estimate)€25,397
- Absenteeism cost avoided (estimate)€0
- CRI at Padoa's price−€6,000
- Net estimated value€19,397
Worked example, conservative scenario: an employer with 1,000 employees pays €6 each, or €6,000 a year. If 6% of results come out Red, 20% of those people act on it and recover 3% of their working time, the estimated productivity value is about €25,397 a year, roughly 4.2 times the cost. The central scenario assumes 10%, 30%, 4% and 1% less absenteeism.2
Inputs from published sources: €43.9 labour cost per hour in France in 20247, 1,607 working hours a year (French legal annual working time) and €1,535 direct cost of absence per employee in 2024.6 Effect sizes are scenario assumptions grounded in workplace trials of sleep programs and resource-building interventions; a pilot would confirm or reject them.2
Seven questions and a timer are easy to copy. A score that means something is not.
A result is only as good as what it is compared with. Padoa would not have to build, norm or validate cognitive assessments itself: that comparison comes from more than twenty years of CogniFit research and data.
210+
peer-reviewed studies
- 3,214,998normed assessments
- 6M+people have used CogniFit
- 18,000+research participants
- 120+universities and research centers
- 25+randomized controlled trials
- 27languages
- Hundreds of thousands of people behind the CRIA normative base built from these same tasks and questions, across ages, education, languages and devices.
- The proprietary CRI modelIts weights were learned from that base and cannot be reverse-engineered from the questions.
- Fair on any deviceEvery session records refresh rate, input method and screen size; corrections come from the same tasks run on thousands of devices.
- Real drops, not bad daysPractice effects and retest noise are known from people who repeated the tasks many times.
- 27 languages, verifiedEvery word is back-translated before release, with the same stimuli everywhere, for multinational workforces.
- Two decades of researchVersioned instruments, traceable data, automated quality checks and peer-reviewed studies.
| What a credible index needs | Build it in-house | Integrate CogniFit |
|---|---|---|
| Validated short tasks and questions | Design and validation studies of its own | Ready |
| Norms by age, sex, education, language and device | Years of collection, with large samples per group | Hundreds of thousands already measured |
| The model linking habits and performance | Not possible until the norms exist | Proprietary, maintained by CogniFit |
| Device and practice-effect corrections | Needs repeated measures at scale | Built in |
| Time to a first pilot | Several years, validity not guaranteed | After the feasibility phase |
These figures describe CogniFit's research and platform as a whole, as published on its developer and scientific validation pages.3 They do not mean that every study validates an occupational outcome, a specific feature or each proposed use case: evidence, endpoints and interpretation rules would be reviewed workflow by workflow with Padoa's medical team.
A specialized cognitive engine, inside the Padoa experience
CogniFit's developer platform gives Padoa a foundation to test this integration without building cognitive assessment, adaptive training and result history from scratch.
Padoa's interface. CogniFit's engine.
Padoa keeps its users, its design and its rules: no cognitive engine to build, no second login.
What CogniFit's developer platform describes
- Server-to-server APIs
- JavaScript, iOS and Android SDKs
- Webhooks and session event callbacks
- Backend-managed account association
- Assessment reports and result history
- White-label and OEM options, under a commercial agreement
1Padoa identifies the user
The authorized user signs in to Padoa, exactly as today.
2A pseudonymous account is associated
Padoa's backend creates or links the matching CogniFit user, server to server, with a pseudonymous ID. The CRI and the full assessments are called with that ID only.
3A short-lived access token
Padoa's backend obtains a short-lived token. Credentials never reach the browser.
4The activity launches inside Padoa
The CogniFit SDK opens the CRI, the assessment or the training in Padoa's web or mobile experience.
5Completion events arrive
Webhooks or session callbacks tell Padoa's backend when a session ends.
6Approved results are retrieved
Padoa's backend retrieves the agreed results and reports.
7Padoa's rules decide who sees what
Clinical, role-based and data governance rules, including the 20-person threshold, apply before anything is displayed.
Padoa experience
Clinician workspace, employee portal, pre-visit questionnaire, administrative workflows.
Secure integration layer
Pseudonymous identity association, server-side credentials, short-lived tokens, activity launch, session events, results retrieval, consent and access control.
CogniFit cognitive technology
CRI engine, assessments, training, skill scores, Cognitive Age where supported, reports and session events.
Role-specific outcomes
Medical information for authorized professionals, personal results for the employee, aggregated program data only where permitted.
Follow one request through the layers
- Launch assessment. The clinician's request in Padoa goes to Padoa's backend, which obtains a short-lived token and opens the CogniFit activity in the employee's Padoa space.
- Assessment completed. CogniFit sends a session event to Padoa's backend. Nothing is shown to anyone yet.
- Retrieve results. Padoa's backend retrieves the approved scores and report, server to server.
- Role permissions. Padoa routes the full result to the employee and the authorized clinician; the employer view receives only aggregated data for groups of 20 or more.
Conceptual simulation: this page makes no network calls. Exact endpoints and events are in CogniFit's developer documentation.
For Padoa's engineers: the SDKs, APIs and event model are documented on CogniFit's developer page.
Designed for trust. Built around medical confidentiality.
Occupational health data is among the most sensitive there is. The concept keeps individual results inside a protected boundary from day one.
Health data, in health hands
Results stay under the occupational health service's responsibility, hosted for health data in the customer's region.
Voluntary, never a condition of employment
People are told plainly what is measured and who sees it, and can see, export and delete their data.
Purpose limitation
Never used to hire, promote, evaluate, rank, discipline or dismiss. It reads no faces, voices or moods.
Human oversight
A qualified professional interprets results and makes every clinical decision.
A real lawful basis
A checkbox alone is not enough inside an employment relationship.
Data minimization
Only what an approved use case needs is collected and transferred, under a pseudonymous ID.
Regulatory review
GDPR, French occupational health rules, medical device rules and the EU AI Act, reviewed for the intended use.
Security documentation
CogniFit's security and privacy materials are in its CogniFit Trust Center.
CogniFit measures and trains cognitive skills; it does not diagnose, treat or prevent any medical condition, and its results are not medical advice and do not replace professional care.
Bring Padoa's medical and privacy leads into the first conversation.
Why habits and a 90-second task describe cognitive resources
The index rests on published research about what drains or builds attention, memory and speed, and on CogniFit's own studies.
The evidence behind the index
- Sleep drives sustained attention (Lim & Dinges, 2010) and work outcomes (Litwiller et al., 2017; Christian & Ellis, 2011).
- Physical activity supports cognition (Erickson et al., 2019; Hillman et al., 2008).
- Social connection has effects on mortality comparable to smoking (Holt-Lunstad et al., 2010).
- Lifestyle builds cognitive reserve: 14 modifiable factors account for about 45% of dementia cases (Livingston et al., 2024; Stern et al., 2020; Schiepers et al., 2018).9
- Self-rated health, a single question, predicts mortality (Idler & Benyamini, 1997).
- Symbol–digit substitution is highly sensitive to cognitive change (Jaeger, 2018), and repeated digital measures are reliable (Sliwinski et al., 2018).
- Cognitive failures predict accidents at work (Wallace & Vodanovich, 2003; Day et al., 2012).
- Resources, not just demands, shape burnout and performance (Demerouti et al., 2001; Bakker & Demerouti, 2007).
- Acting on resources can pay. Digital CBT for insomnia at work improved sleep and productivity (Bostock et al., 2016; Espie et al., 2019); workplace resources improve well-being and performance (Nielsen et al., 2017); organizational actions outperform individual-only ones (LaMontagne et al., 2007).
Measure to act, not just to measure
Wellness programs that only measure have shown little effect (Song & Baicker, 2019; Jones et al., 2019; Fleming, 2024). The CRI is built to trigger targeted action, and the pilot is designed to test whether it does.
CogniFit's track record
- Chronic insomnia and cognitive functioning in older adults (Haimov, Hanuka & Horowitz, 2008).
- Cognitive training and sleep quality in older adults with insomnia (Haimov & Shatil, 2013).
- Combined cognitive and physical training (Shatil, 2013).
- Personalized cognitive training versus computer games, a randomized double-blind trial (Peretz et al., 2011).
- Online cognitive training at home with older adults (Gigler et al., 2013).
- Cognitive training in intellectual and developmental disabilities (Siberski et al., 2014).
Who works with CogniFit today
- Healthcare professionals and clinics, who use its assessments to support cognitive evaluation and follow-up
- Researchers, universities and research centers, running studies with its tasks
- Organizations offering cognitive wellbeing programs to their people
- Educators and schools
- Individuals and families, directly on CogniFit
Named client references and case studies can be discussed in the first meeting, with each organization's permission.
TIME Best Inventions of 2026. TIME included CogniFit's Digital Cognitive Twin in its list. The Digital Cognitive Twin is a separate, future concept, not part of the APIs in this proposal, and the recognition implies no endorsement of this partnership by TIME.
References for the evidence above
- Bakker & Demerouti (2007). Journal of Managerial Psychology. doi.org/10.1108/02683940710733115
- Bostock, Luik & Espie (2016). Journal of Occupational and Environmental Medicine. doi.org/10.1097/JOM.0000000000000778
- Christian & Ellis (2011). Academy of Management Journal. doi.org/10.5465/amj.2010.0179
- Day et al. (2012). Accident Analysis & Prevention. doi.org/10.1016/j.aap.2012.03.028
- Demerouti et al. (2001). Journal of Applied Psychology. doi.org/10.1037/0021-9010.86.3.499
- Erickson et al. (2019). Medicine & Science in Sports & Exercise. doi.org/10.1249/MSS.0000000000001936
- Espie et al. (2019). JAMA Psychiatry. doi.org/10.1001/jamapsychiatry.2018.2745
- Fleming (2024). Industrial Relations Journal. doi.org/10.1111/irj.12418
- Gigler et al. (2013). Gerontechnology. doi.org/10.4017/gt.2013.12.1.007.00
- Haimov, Hanuka & Horowitz (2008). Behavioral Sleep Medicine. doi.org/10.1080/15402000701796080
- Haimov & Shatil (2013). Sleep Medicine. doi.org/10.1016/j.sleep.2013.11.053
- Hillman, Erickson & Kramer (2008). Nature Reviews Neuroscience. doi.org/10.1038/nrn2298
- Holt-Lunstad et al. (2010). PLoS Medicine. doi.org/10.1371/journal.pmed.1000316
- Idler & Benyamini (1997). Journal of Health and Social Behavior. doi.org/10.2307/2955359
- Jaeger (2018). Journal of Clinical Psychopharmacology. doi.org/10.1097/JCP.0000000000000941
- Jones, Molitor & Reif (2019). Quarterly Journal of Economics. doi.org/10.1093/qje/qjz023
- LaMontagne et al. (2007). International Journal of Occupational and Environmental Health. doi.org/10.1179/oeh.2007.13.3.268
- Lim & Dinges (2010). Psychological Bulletin. doi.org/10.1037/a0018883
- Litwiller et al. (2017). Journal of Applied Psychology. doi.org/10.1037/apl0000169
- Livingston et al. (2024). The Lancet. doi.org/10.1016/S0140-6736(24)01296-0
- Nielsen et al. (2017). Work & Stress. doi.org/10.1080/02678373.2017.1304463
- Peretz et al. (2011). Neuroepidemiology. doi.org/10.1159/000323950
- Schiepers et al. (2018). International Journal of Geriatric Psychiatry. doi.org/10.1002/gps.4700
- Shatil (2013). Frontiers in Aging Neuroscience. doi.org/10.3389/fnagi.2013.00008
- Siberski et al. (2014). American Journal of Alzheimer's Disease & Other Dementias. doi.org/10.1177/1533317514539376
- Sliwinski et al. (2018). Assessment. doi.org/10.1177/1073191116643164
- Song & Baicker (2019). JAMA. doi.org/10.1001/jama.2019.3307
- Stern et al. (2020). Alzheimer's & Dementia. doi.org/10.1016/j.jalz.2018.07.219
- Wallace & Vodanovich (2003). Journal of Occupational Health Psychology. doi.org/10.1037/1076-8998.8.4.316
Start focused. Validate demand. Build only what creates value.
One 12-month pilot with one occupational health service and 2,000 to 5,000 employees, run with Padoa's product, medical and commercial leadership, and a clear go or no-go at the end.
- 1Kick-off
Joint opportunity validation
- Choose the service and the employers taking part
- Identify the economic buyer
- Interview occupational health stakeholders
- Review workflows and legal constraints
- Agree success criteria and targets
- 2Weeks 1–8, target
Integration feasibility
- Validate API and SDK fit
- Map pseudonymous identity and permissions
- Define data governance
- Put the CRI in the pre-visit questionnaire
- Test clinician and employee experiences
- 3Quarters 1–4
Controlled customer pilot
- Personal results, clinician view and team barometer
- A limited, authorized population
- Activation, engagement and workflow effort
- Professional feedback
- Willingness to pay
- 4Month 12
Report and commercial decision
- Proceed to a product
- Change the scope and rerun
- Limit the use case
- Decline the opportunity
How the pilot would be judged
- Customer demand
- Workflow and integration feasibility
- Completion and engagement
- Clinical and user acceptability
- Commercial viability
Success metrics
| Metric | What it shows | Target |
|---|---|---|
| Test-retest reliability of the CRI | Whether the index is stable when nothing has changed | To be agreed |
| Participation | How many invited employees take the CRI each quarter | To be agreed |
| Red results that lead to a consultation | Whether the signal reaches the occupational health team | To be agreed |
| Professional satisfaction | Whether clinicians find the view useful in their workflow | To be agreed |
| Employer willingness to pay | Whether employers would buy it at Padoa's price | To be agreed |
| Effect on absenteeism | Whether teams that act see a change, compared over the year | To be agreed |
A focused pilot is the fastest responsible way to find out whether this should become a Padoa product line.
Questions Padoa's teams are likely to ask
Want to scope one use case?
A first call with Padoa product leadership, an occupational health representative and CogniFit leadership.
Would employers see an employee's cognitive results?
No. Individual results stay with the employee and authorized occupational health professionals. Employers would see program participation and, only where a lawful basis exists, a team barometer for groups of 20 people or more; smaller groups stay hidden.
Is the CRI a diagnosis or a medical device?
It is not a diagnosis. The CRI is a prevention indicator of cognitive resources and habits; it does not assess fitness for work. Whether a given use falls under medical device rules depends on its intended use, and that would be reviewed with legal and clinical teams before any pilot.
Why is a 90-second task enough?
It is enough for a short indicator, not for a full assessment. Symbol–digit tasks are sensitive to changes in speed and attention, and repeated digital measures are reliable; combined with seven habit questions, norms and a personal baseline, they give a useful signal. When the signal or the clinician calls for more, the full Cognitive Check takes about 20 minutes.
What if someone answers at random or repeats it many times?
Quality control flags rushed or interrupted sessions instead of scoring them. Practice effects and retest noise are known from people who repeated the tasks many times, so the personal baseline discounts them, and the program schedules one check per quarter.
What exactly does the employer see, and from how many people?
Team shares of Green, Amber and Red, their trend, the main drivers and suggested actions, plus program participation. Only for groups of 20 people or more, only where a legal basis exists, and only once the occupational health service has validated the report. Never a name, an individual score or a ranking.
Who invoices whom?
In the proposed model, employers or occupational health services pay Padoa at prices Padoa sets. CogniFit invoices Padoa: €3 per enrolled employee per year for the CRI with unlimited checks, and each full assessment on a cost that falls with volume, as used in the simulator. Final terms would come out of the commercial discussion.
Does CogniFit diagnose conditions or certify fitness for work?
No. CogniFit measures and trains cognitive skills. Its results support a professional's judgment; they do not diagnose, establish fitness for work or replace clinical care.
Is there a CogniFit API or SDK Padoa could use?
Yes. CogniFit's developer platform describes server-to-server APIs, JavaScript, iOS and Android SDKs, webhooks and session callbacks, and assessment reports. Which pieces fit Padoa is what the feasibility phase would confirm.
Would employees need a separate CogniFit login?
No, in the proposed design. Padoa's backend would associate each authorized user with a pseudonymous CogniFit account, so people stay inside Padoa.
Could it be offered as a Padoa-branded product?
Possibly. CogniFit offers white-label and OEM options under a commercial agreement; the scope would be part of the commercial discussion.
Is the revenue simulator a forecast?
No. It multiplies the figures you enter and subtracts the scenario CogniFit costs. Final pricing, distribution rights, revenue share and other costs would come from commercial due diligence.
How much would the integration cost, and how long would it take?
That is not known yet. The pilot plan sets an eight-week target for the connection; the feasibility phase would confirm it, along with the effort and costs. CogniFit does not look to earn its margin on the integration work itself, but on the long-term usage volume that a successful partnership would generate, so both sides gain from keeping the integration lean.
Would this compete with Padoa's mental health and psychosocial prevention offers?
No. The intent is a complementary measurement and training layer. Whether a real product gap exists is something to validate with Padoa's teams.
Is the Digital Cognitive Twin part of this proposal?
No. It is a separate CogniFit concept. Any role for it would be a later research conversation with its own scope, validation and regulatory review.
Which regulations would apply?
It depends on the intended use, the implementation and the country. GDPR, French occupational health requirements, medical device rules and the EU AI Act would all be reviewed by legal and clinical teams before any pilot.
What if cognitive health became a new pillar of Padoa's prevention platform?
Padoa has the occupational health workflows and the reach. CogniFit has a research-based index, full assessments and training, available through integration tools. The next step is simple: decide whether a 12-month pilot with one occupational health service is worth running.
Proposed first discussion
Padoa product leadership, an occupational health representative and CogniFit leadership.
- Strategic fit
- Priority use case
- Integration feasibility
- Commercial model
- Pilot scope
Proposed introduction, ready to forward
CogniFit has prepared a concept showing how Padoa could add cognitive health in three layers: a 3-minute Cognitive Resources Index for every employee, a full CogniFit assessment when the index or the clinician calls for it, and personalized training with team actions, all built on CogniFit's APIs and SDKs. It is a hypothesis, not an agreement, with synthetic demo data and editable scenarios. Could we set up a short call between Padoa's product and medical leadership and CogniFit to scope a 12-month pilot with one occupational health service?
Request a pilot scoping meeting
Tell us who should join from Padoa and which use case matters most. The form prepares an email to CogniFit's sales team at [email protected] in your email app, so nothing leaves your device until you press send.
- Who should join from product, medical or commercial leadership
- The use case and the occupational health service you would start with
- Any constraint we should know before the call
Prefer to write directly? The address is [email protected].
Notes and sources
This page is an independent partnership concept prepared by CogniFit. It does not represent an announced partnership, a product commitment or an endorsement by Padoa or Thoma Bravo. All personas, scores, program figures and revenue scenarios on this page are synthetic or hypothetical, and proposed Padoa features are not existing Padoa functionality.
- CogniFit. Company website and product positioning: digital cognitive assessment and personalized cognitive training for individuals, healthcare professionals, researchers and organizations, used by more than 6 million people. cognifit.com
- CogniFit. Padoa Cognitive Health partnership concept (2026): synthetic demonstration data for the clinician, employee and employer views, example funnel shares, and hypothetical assumptions for the scenarios, including the €3 CRI price to Padoa, the per-assessment cost curve and the employer effect sizes, prepared for this page. They are not measured outcomes, quotes or forecasts.
- CogniFit. Developer platform and scientific validation pages: research figures (210+ peer-reviewed studies, 18,000+ research participants, 120+ universities and research centers, 25+ randomized controlled trials), 3,214,998 normed assessments, 27 languages, SOC 2 and GDPR. cognifit.com/developers · cognifit.com/scientific-validation
- TIME. Best Inventions of 2026: CogniFit Digital Cognitive Twin.
- French occupational health and prevention services (SPSTI). Annual activity figures, 2024: mid-career visits.
- WTW. Absenteeism barometer, France, 2024: private-sector absenteeism rate and direct cost of absence per employee.
- INSEE. Hourly labour cost in France, 2024.
- Hafner M, Stepanek M, Taylor J, Troxel WM, van Stolk C (2017). Why sleep matters: the economic costs of insufficient sleep. RAND Corporation. doi.org/10.7249/RB9962
- Livingston G, et al. (2024). Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet. doi.org/10.1016/S0140-6736(24)01296-0
United States