A 30-day, read-only Snapshot of up to 10 commercial rooftops across two monitoring portals: which sites stopped reporting, which are stale, which never made it into the portal at all — and how old each gap was when the window closed.
For O&M operations leads and installers responsible for 20–200 commercial rooftops. Delivered remotely from Poland, in writing.
Founding cohort · first five Snapshots at a fixed price
Synthetic example · Not customer data↳
01 / Portal overview6 sites · 0 alerts
✓✓✓✓✓✓
Snapshot: site list vs. records actually received↓
02 / Sites fully reporting3 of 7
✓?✓!—✓!
!
In this example, four blind spots the overview did not flag: one silent for 11 days, one stale, one never mapped into a portal, and one site still reporting with an inverter absent.
Illustration from the synthetic example below: the portal overview can list only 6 of the 7 sites in the portfolio. The rest were never mapped into a portal at all.
At handover
A portfolio changes hands faster than its monitoring does. Portal accounts, viewer roles and site mappings can lag behind the sites themselves, and a site that was never mapped into a portal does not appear in that portal's overview at all. An overview alone does not establish that every expected site and device is mapped and supplying complete records; comparing it with the expected register does. The Snapshot reconciles the site list you are signing for against the records the portals actually received in your window. It reads historical exports; it does not verify a portal migration, an API change or future continuity.
THE SNAPSHOT
Ten sites. Two portals. Thirty days. One list you can act on.
Portfolio Blind-Spot Snapshot
I reconcile your site list against the records your monitoring portals actually received in one 30-day window. You get an exception list per site with evidence — not a dashboard, and not a promise to watch it for you.
Fixed scope, fixed price, read-only. Written report, one written Q&A round, and one correction if I got something wrong on the inputs we agreed.
Scope
10 sites · 2 portals · 30 days
One window you choose — for example the month before handover.
Turnaround
10 working days
Ten working days start once complete inputs and a reserved slot are agreed. The scope is fixed; if your data needs more than it covers, I say so before starting.
Price
£450fixed for UK customers, in pounds sterling · founding cohort of five
Invoice from SHEV Software (Poland). Payment terms are agreed in writing before the work starts.
You provide
A site list and the exports you already have
Site list with device IDs and expected reporting cadence
The CSV/XLSX exports you already download — or, if that is not possible, a temporary viewer role
You receive
An exception list per site
Reporting · stale · silent · missing from the portal
First and last record in the window
Gap age at the review cutoff: how long each gap had been open when the window closed
Evidence for every finding, reproducible
Not included
No visits. No diagnosis. No monitoring.
Site visits, equipment or electrical diagnosis
Yield or performance guarantees
Continuous monitoring, alerts, control of any system
I tell you which sites are silent — not why.
After the Snapshot: if the list is useful, the next step is a full Fleet Blind-Spot Audit — 20–200 sites, up to three portals — scoped and priced from what the Snapshot found.
WHAT A FINDING LOOKS LIKE
Silent is not the same as zero.
Synthetic example · Not customer data
Seven fictional rooftops, two portals, the 30 days before a handover. The portal overview showed no alerts. The Snapshot compares what each site was expected to send with what was actually received.
Fictional portfolio. Window: 25 September – 24 October 2026, 30 days. Expected cadence is a rule of this invented example, not an OEM claim. Site names and portals are placeholders. Gap age is measured to the last day of the window; in this example it is a declared assumption of the fixture, not evidence that anyone noticed anything.
Site
Portal
Last record in window · UTC
Expected
Snapshot status
Gap age at cutoff
S-01 Warehouse North
Portal A
2026-10-24 17:45Z
15 min
Reporting
—
S-02 Logistics Hub
Portal A
2026-10-13 09:15Z
15 min
Silent
11 days
S-03 Retail Park B
Portal B
2026-10-24 17:30Z
15 min
Reporting
—
S-04 Cold Store
Portal B
2026-10-21 12:00Z
15 min
Stale · partial days since 2026-10-19
5 days
S-05 Office Roof
— Not in any portal
—
daily
Missing from both portals
whole window
S-06 Depot East
Portal A
2026-10-24 17:45Z
15 min
Reporting
—
S-07 Plant Two
Portal A
2026-10-24 17:45Z
15 min
Reporting · inverter 3 of 4 absent since 2026-10-19
5 days
In this example S-05 is the case the overview does not show: a site that was commissioned but never mapped into a portal is not in that portal's overview at all. S-07 is the other: the portal still reports it while one of its four inverters has been absent since 19 October. Whether a particular portal would raise an alert for that is a property of that portal, which this fictional example does not establish.
↳
Four sites need a decision before sign-off.
The overview counted devices that exist. The Snapshot counts records that arrived. In this example 4 of the 7 sites carry a gap that nothing surfaced during the window — including one the portal still lists as reporting, because only one of its four inverters stopped arriving.
The Snapshot cannot establish why a site is silent — communications, inverter, portal mapping — or what the lost yield is worth. It establishes which sites, since when, and what was visible meanwhile. Gap age is measured to the review cutoff; it is not a measurement of when anyone noticed.
You send the number of sites, the portals, and whether you can share exports or a viewer role. My first written reply comes within two working days — that is a reply, not a decision on your data. Before anything is committed I check a sample export for suitability, then confirm the window, the reserved slot and the fixed price.
02
Read-only reconciliation
You create a temporary viewer role or share exports. I compare site list, expected cadence and received records, portal by portal. Nothing is changed and nothing is controlled.
03
Exception list with limits
You receive the per-site list, the evidence behind each finding, and an explicit list of what the data could not show. Access is revoked and exports are deleted 30 days after delivery.
WHO DOES THE WORK
An engineer with no field crew to sell you.
I am Oleksii Shevchenko, a senior Go backend engineer — distributed systems, data pipelines, workflow orchestration — based in Poland. Enerplain is the energy-data line of my company, SHEV Software. I do every Snapshot myself, in writing, in English (also Ukrainian and Russian).
The Snapshot has nothing to gain from finding faults: no maintenance contract follows and no hardware is sold. Wind and storage portfolios are a later interest, not a current service.
A few useful boundaries.
Is this monitoring?+
No. One 30-day window, checked once, delivered once. No dashboard, no alerts, no control of any system. If you want it repeated, that is a separate conversation after the first list.
Which portals can you work with?+
There is no universal portal compatibility, and I claim no verified support for any specific OEM. It works where a portal offers a viewer role or a CSV/XLSX export carrying site, timestamp and energy columns — I check a sample of your actual export before committing. The first Snapshot covers two portals; a mixed estate with more goes into the full audit.
Do you need our credentials?+
No. A temporary read-only role created for the Snapshot and revoked afterwards, or exports. Never send passwords, customer identifiers or raw telemetry by email — the first email needs only counts and portal names. Once the scope is agreed, what I need is the site and device register, the exports themselves, and which portal and export produced each file.
What can the Snapshot not tell us?+
Why a site is silent, whether equipment is faulty, or the value of lost yield. This is not electrical diagnosis, performance certification or energy-savings verification. A gap is an evidence gap: records did not arrive under the cadence we agreed. It does not by itself mean a fault — a communications link, a data logger, a portal mapping or the export itself can produce the same silence.
FIELD NOTES
The details that make a monitoring export lie.
Short, practical notes for people who read PV data, not just dashboards. Current notes are in English.
How many sites, which portals, and is a viewer role possible?
That is the whole first email. My first written reply comes within two working days — a reply, not yet a commitment. No attachments, credentials or customer data are needed.
Your mail app will open. Correspondence in English; Ukrainian and Russian on request. Delivered remotely from Poland. Please do not attach customer data or credentials.