ANNAPURNA AGENTIC SOLUTIONSSURYAPULSE
Research pilot · Invitation only
Solar monitoring · Evidence assurance

One rooftop. Fourteen days. A clearer view of solar data.

We are seeking one authorized commercial or institutional rooftop solar installation for a bounded, read-only research pilot. The aim is to assess data quality, available inverter exports, irradiance evidence and independent operator review—not to promise fault detection or financial savings.

No data-submission form, tracking cookies, analytics or physical actuation. Pilot eligibility is assessed before any transfer.

Evaluation design

What a pilot actually includes

A tightly scoped, reproducible process that is useful even if the answer is “the measurements are not yet reliable enough for diagnostics.”

01 / OBSERVE

Check the source

Verify timestamp cadence, inverter-energy units, irradiance provenance, missing intervals and calibration records.

02 / REVIEW

Evaluate evidence

Freeze candidate findings, then ask a separate qualified reviewer to adjudicate real positive and negative events.

03 / DECIDE

Report what is known

Deliver a redacted quality and usefulness report. Where inputs are inadequate, the report says so explicitly.

Eligibility and safety

What the site must provide

Minimum technical inputs

  • 14 days of time-aligned solar inverter energy exports
  • Plane-of-array irradiance observations for performance claims
  • PV site configuration and known timestamp/interval units
  • Instrument type, calibration and uncertainty evidence
  • An operator or reviewer independent of the model author

Site ownership and safeguards

  • Written authority from the actual data/controller owner
  • Agreed purpose, allowed fields, retention and deletion terms
  • Read-only exports; no equipment commands or modification
  • No customer identity or sensor upload to this page
  • Separate approval before any external publication
Research, not certification. This prototype is not IEC-certified, does not claim production-readiness, and cannot substantiate fault rates, site performance or savings without independently validated field evidence. Without calibrated POA data, the study will be restricted to data-quality findings.
Four-step onboarding

From introduction to review

01 · Discovery

Discuss the site and available export types; no raw data yet.

02 · Authorization

Confirm owner approval, rights, scope, retention and reviewer.

03 · Intake

Collect the agreed 14-day exports in a protected local workflow.

04 · Evidence

Review incidents and publish only approved, caveated outputs.