Search all 478 artefacts by title, document ID or content.
Printed on the standard letterhead. Page furniture, margins and repeating table headers come from the same stylesheet the PDF service uses.
Pensieve Labs
The operating system for hospitals
STM-GL-026
v1.0.0 | 01 August 2026
STM-GL-026 | Version 1.0.0 | Last Modified On 01 August 2026
Any supplier can publish an availability target. This Statement publishes the achieved figure, month by month, including the months it was missed, and states the rules that stop the archive from being quietly improved later.
The live status page is at Pensieve status page URL. This Statement is the archive and the rules;
the status page is the live instrument. Neither restates the measurement methodology, which is
SLA-GL-001 clause 4 and is the source of truth.
The governing commitment: a month that missed its target is published as a month that missed its target, in the same place and the same format as every other month, and is never removed.
| Deployment model | What is published |
|---|---|
DM-1 Dedicated |
Monthly availability per deployment, and the aggregate across DM-1 deployments |
DM-2 Shared |
Monthly availability of the shared platform, by component |
DM-3 Customer Cloud |
Availability of the Pensieve-operated components only. Infrastructure faults in the hospital's own cloud account are the hospital's to measure, and are excluded, with the exclusion shown, not hidden |
DM-4 On-Premise |
Not measured by Edsol Edtech Pvt. Ltd.. Pensieve does not operate the infrastructure and does not publish a figure it cannot measure. SLA-GL-001 clause 6 states what applies instead |
| Element | Detail |
|---|---|
| Component status | Per component: application, database, integrations, authentication, background processing, reporting |
| Current incidents | Opened within fifteen (15) minutes of confirmation, with what is affected and what is not |
| Incident updates | At least every sixty (60) minutes while a Severity 1 is open, matching the ticket cadence in POL-GL-056 4.5 |
| Scheduled maintenance | Posted in advance, per the notice periods in SLA-GL-001 clause 10 |
| Historical uptime | Rolling ninety-day view, plus the monthly archive at 3 |
| Subscription | Anyone may subscribe, customer or not, without an account |
1.1 The status page is not gated. A prospective customer, a journalist and a competitor see the same page a customer does. A status page visible only to customers is a marketing page.
1.2 A component is marked degraded when it is degraded, not when the ticket volume becomes uncomfortable. The decision to post is made by the duty engineer on the technical signal, and no commercial approval is required or sought.
2.1 No retrospective editing. A published incident record is never deleted, and its timeline is never altered. A correction is published as a correction, dated, with the original text retained.
2.2 No quiet re-classification. An incident's severity is not lowered after the fact to improve a figure. Where a severity is corrected, the correction and its reason are shown.
2.3 Missed months stay published. Availability below target is published with the same prominence as availability above it.
2.4 Exclusions are shown, not applied silently. Where an excluded window under SLA-GL-001 clause 4 has
removed time from the calculation, the archive shows both the raw figure and the figure after exclusions,
with the exclusion category. A methodology that only ever produces flattering numbers is not a methodology.
2.5 The archive is permanent. Entries are retained for the life of Edsol Edtech Pvt. Ltd., and the
archive is included in the exit pack under WPR-GL-400 and in the wind-down obligations in DIS-GL-407.
2.6 Post-incident reviews are published for every Severity 1, within ten (10) Business Days, at the tier stated at 4.
| Period | DM-1 achieved |
DM-2 achieved |
DM-3 achieved |
Target | Raw before exclusions | Excluded time and category | Severity 1 incidents | Service Credits accrued |
|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The rendered archive carries one row per completed month since Pensieve status page first month. The
three rows above are the template shape.
3.1 Basis of measurement: SLA-GL-001 clause 4. Measurement source: the monitoring described in
POL-GL-113, extracted on Uptime extraction date. Independently verified: no, see
5.
For each incident, published at the Public tier:
| Field | Content |
|---|---|
| Reference | |
| Component and deployment models affected | |
| Detected at / resolved at | |
| User-visible effect | |
| What was not affected | |
| Cause, in plain language | |
| What was changed so it does not recur, and by when | |
4.1 What is never published in an incident record: exploitable technical detail, credentials or
identifiers, any customer's name without consent under ADD-GL-020, and any personal data. Where an
incident involved personal data, the notification obligations in DIS-GL-016 and DPA-GL-001 operate
separately and directly to the affected customers; a status page post is never a substitute for a breach
notification.
4.2 The "what was not affected" field is deliberate. During an incident, most of a hospital's anxiety is about scope. Stating what is unaffected is more useful than another sentence about what is.
5.1 Operating history is short. Edsol Edtech Pvt. Ltd. is early, and a small number of months produces
a figure that is arithmetically true and statistically weak. A single incident moves a monthly percentage a
long way. Read the incident records, not only the percentage.
5.2 No third party verifies these figures. They are produced from Pensieve's own monitoring. Where an independent assessment is later obtained, its scope will be stated here.
5.3 Monitoring measures what it is pointed at. Availability is measured by the synthetic checks and
component probes described in POL-GL-113. A failure mode nobody has thought to probe is a failure mode
that would not appear in these figures until a hospital reports it. This is true of every status page ever
published, and most do not say so.
5.4 DM-4 is absent, not zero. See the applicability table.
5.5 Per-hospital figures are not published. A hospital's own dedicated deployment figure is in its
Service Report under SLA-GL-001 clause 11. Publishing per-hospital availability would identify the
hospital.
| # | Testable statement | Evidence |
|---|---|---|
| T-1 | Every Severity 1 incident has a status page post opened within 15 minutes of confirmation | REG-GL-203 against status page timestamps |
| T-2 | Every Severity 1 has a published post-incident review within 10 Business Days | Incident register against publications |
| T-3 | No published incident record was deleted or its timeline altered | Archive integrity check, with content hashes |
| T-4 | Every month since the first is present in the archive, including months below target | Completeness check against the calendar |
| T-5 | Raw and post-exclusion figures are both published for every month with an exclusion | Archive review |
| T-6 | No incident record contains personal data, a customer name without consent, or exploitable detail | Pre-publication review record |
| T-7 | The archive figure for a month matches the Service Reports issued for that month | Reconciliation, quarterly |
| Token | Meaning |
|---|---|
Pensieve status page first month |
The first month for which an availability figure exists |
Uptime extraction date |
Date on which the archive figures were extracted from monitoring |
| Document | Relationship |
|---|---|
SLA-GL-001 clause 4 and clause 11 |
Source of truth for measurement, exclusions and Service Reports |
DIS-GL-030 Uptime, Performance and Capacity Disclosure |
What the Platform is engineered to sustain |
POL-GL-113 Logging and Monitoring Policy |
How availability is measured |
POL-GL-112 Incident Response Policy, REG-GL-203 Incident Register |
Where incidents are managed and recorded |
DIS-GL-016 Incident Response and Breach Notification |
Why a status post is never a breach notification |
POL-GL-056 Support Policy and Escalation Matrix |
The update cadence this Statement matches |
NTC-GL-005 Incident and Maintenance Notices |
The notice instruments |
POL-GL-068 Transparency Report |
Where the semi-annual summary appears |
WPR-GL-400, DIS-GL-407 |
Why the archive survives an exit or a wind-down |
| Version | Date | Author | Summary |
|---|---|---|---|
| 1.0.0 | 2026-08-01 | Engineering | First publication. Establishes an ungated status page, a fifteen-minute posting rule and a sixty-minute update cadence, and five archive-integrity rules: no retrospective editing, no quiet re-classification, missed months stay published, exclusions shown alongside raw figures, permanent retention. Publishes the monthly archive and incident record shapes, requires a "what was not affected" field, and states honestly that operating history is short, that no third party verifies the figures, and that monitoring only measures what it probes. |
STM-GL-026 v1.0.0 | Last Modified On 01 August 2026 | Review due
31 January 2027 | Published at https://trust.pensievelabs.org and Pensieve status page URL