All features

Stages 1–5 · Foundation

PV modeling, weather & data foundation

Application structure, organizations/sites, reference PV physics, and historical weather support the product behind the scenes.

Embedded infrastructure, not five separate customer screens. Weather and model inputs must be usable.

Stage 6 · Site analytics

Weather-adjusted expected production

Compare observed AC production with the site's modeled expectation.

Requires site configuration, aligned telemetry, and suitable weather inputs.

Stage 7 · Foundation

Synthetic fault validation

Deterministic fault fixtures support engineering regression checks.

Internal validation tooling, not customer measurements or evidence of field accuracy.

Stage 8 · Data & investigations

Fault detection & data-quality guardrails

Inspect outage, underperformance, clipping, soiling-like, and communications evidence.

Detection is evidence-dependent. Missing telemetry is not zero production.

Stage 9 · Data & investigations

Lost energy, value & priority

Review interval-integrated loss estimates and investigation priorities.

Financial impact requires configured energy value. Estimates are not guaranteed savings.

Stage 10 · Data & investigations

CSV ingestion & normalization

Import measured CSV history with explicit column mapping, units, timezone, and quality warnings.

Authorized site only; matching timestamps may be updated after confirmation. Keep the original export: hosted raw copies are temporary.

Stage 11 · Data & investigations

Persistent telemetry & history

Inspect stored telemetry coverage and daily production history.

Persistence and import processing are backend services; a report does not create live telemetry.

Stage 12 · Foundation

Tenant-scoped analytics APIs

Validated APIs power the authorized portfolio and site views.

Infrastructure behind the product. Site identifiers do not grant access to another organization.

Stage 13 · Operations & account

Fleet dashboard & action queue

Compare sites, filter the portfolio, review losses, and prioritize investigation.

Customer totals need analyzed site data. Sample totals are independent fictional examples.

Stage 14 · Site analytics

Site analysis & production history

Inspect actual versus expected production, quality, history, and analysis findings.

Choose a site; each analysis keeps its own evidence and readiness state.

Stage 15 · Data & investigations

Diagnostics & operational findings

Review likely causes, supporting and conflicting evidence, and incident context.

A probable cause is not confirmed by a model score.

Stage 16 · Operations & account

Portfolio reports & delivery schedules

Review printable portfolio summaries and manage opt-in Site Analysis PDF schedules.

Scheduled core PDFs require eligible plan, site data, opt-in, and configured delivery. Printing a portfolio page is a separate action.

Stage 17 · Operations & account

Team, roles & account security

Manage organization membership and account access using existing permissions.

Privileged team changes remain owner/admin-only. The public demo cannot impersonate an account.

Stage 18 · Foundation

Deployment & service health

Runtime health and operational infrastructure support the customer application.

Health endpoints are not proof that every provider, model, or customer workflow is ready.

Stage 19 · Foundation

Private real-world validation

Private, Git-ignored validation workflows evaluate eligible operational datasets.

Operator tooling, not a public customer-data browser. Synthetic tests do not establish field validation.

Stage 20 · Operations & account

Subscriptions & billing

Review your plan, site allowance, and authorized billing actions.

Hosted billing and entitlement checks remain separate from scientific model readiness.

Stage 21 · Data & investigations

Site imports & monitoring integrations

Import supported site designs and configure optional provider connections.

Provider registration, credentials, mapping, and capability qualification are required; no demo connection is active.

Stage 22 · Site analytics

Inverter-level monitoring

Compare device output, peers, outages, and equipment evidence.

Requires suitable device telemetry and plan/provider capability. Site-level data does not prove inverter health.

Stage 23 · Site analytics

String & MPPT monitoring

Drill into DC channel power, current, voltage, and comparable siblings.

Open a device, then its channel. Requires explicit topology and DC measurements.

Stage 24 · Data & investigations

Advanced fault classification

Inspect differentiated equipment, environmental, and data-related evidence.

Classification keeps uncertainty and alternative explanations; it does not confirm a physical failure.

Stage 25 · Foundation

ML-assisted classification

The model registry, shadow evaluation, fallback, and activation gates support classification.

A visible feature does not make the model ACTIVE. Evidence and authorized promotion are still required.

Stage 26 · Site analytics

Degradation analysis

Review normalized long-term trends and history limitations.

Short-history estimates remain preliminary, not reliable long-term degradation proof.

Stage 27 · Site analytics

Soiling & cleaning

Inspect accumulation, recoveries, method agreement, uncertainty, and cleaning economics.

Requires usable performance history; rainfall alone does not prove a clean array.

Stage 28 · Site analytics

Snow coverage & loss

Review weather and electrical snow evidence, modeled coverage, recovery, and energy loss.

Missing telemetry cannot confirm snow loss; weather-only estimates remain labeled.

Stage 29 · Site analytics

Shading & shading change

Compare solar-position fingerprints, changes, affected areas, and supported impact.

Requires comparable baseline coverage; no automatic obstruction or vegetation confirmation.

Stage 30 · Site analytics

Curtailment & active-power limits

Separate binding control limits, clipping, and additional performance deficit.

Analytics-only. No plant commands or setpoint writes are enabled.

Stage 31 · Site analytics

Availability

Inspect equipment state, outages, unknown time, capacity weighting, and policy results.

Contract claims require a valid policy. Communications failure is not a confirmed equipment outage.

Stage 32 · Site analytics

Performance ratio & KPIs

Review PR, yield, coverage, boundaries, methodology, and losses.

Requires compatible irradiation, capacity, energy, and measurement boundaries.

Stage 33 · Site analytics

Energy & power forecasting

Inspect forecast power, uncertainty, freshness, assumptions, and accuracy.

Weather vintages and operational trials govern evaluation; SHADOW correction does not imply active hybrid forecasting.

Stage 34 · Site analytics

Predictive maintenance

Review component health, risk horizons, evidence, confidence, and impact.

Predictive model activation requires field evidence. Missing data and cold starts retain fallback/abstention.

Stage 35 · Site analytics

Maintenance recommendations

Compare actions, do-nothing consequences, economics, constraints, and timing.

Human approval is required. SHADOW recommendations do not authorize customer action or dispatch.

Stage 36 · Operations & account

Work-order management

Review drafts, approval, scheduling, checklists, execution, verification, and history.

Commands require permitted roles and audit reasons; completion and verified recovery are distinct.

Stage 37 · Operations & account

CMMS integration

Manage authorized MaintainX connection, mappings, export, synchronization, and conflicts.

Needs configured credentials/provider access and approved export. No automatic public-demo dispatch.

Stage 38 · Operations & account

Alerting & notifications

Inspect deduplicated episodes, grouping, suppression, routing, and delivery history.

External delivery requires verified destinations, configuration, and eligible policy; the demo sends nothing.

Stage 39 · Site analytics

Portfolio benchmarking

Compare technically compatible peers and inspect qualified performance opportunities.

A published run and sufficient comparable peers are required; no results is not zero performance.

Supporting workflow · Data & investigations

Historical forensic investigation

Upload, map, quality-check, and analyze an operational CSV without a live SCADA connection.

Authorized customer uploads only. Public previews neither receive a file nor run an analysis.

Supporting workflow · Data & investigations

On-demand Site Analysis PDF & email

Generate or email a report from the current completed Site Analysis.

Run Site Analysis first. Email requires an authorized recipient and configured delivery; this is distinct from portfolio printing.

Supporting workflow · Data & investigations

Investigation evidence packets

Open a reproducible report from a persisted finding, with an anonymized version and PDF.

Run and save an analysis, then select Open investigation evidence packet on a persisted finding.

Supporting workflow · Data & investigations

Resolution feedback & learning loop

Record field outcomes while retaining original predictions and append-only history.

Record investigation outcome is offered only for persisted customer findings. No demo confirmation is saved.

Supporting workflow · Data & investigations

Next diagnostic actions & decision economics

Compare checks, required evidence, information gain, costs, and conditional recoverable value.

Decision support, not repair instructions. Cost inputs and expected recovery assumptions remain explicit.

Supporting workflow · Data & investigations

Diagnostic performance

Review outcome-based precision, calibration, and recurring confirmed causes.

Owner/admin access and resolved labels are required. No field labels means no verified accuracy.

Supporting workflow · Operations & account

Fleet pattern intelligence

Review persistent patterns across genuinely comparable site cohorts.

Requires sufficient comparable and independent reference sites; correlations do not prove shared causes.

Supporting workflow · Operations & account

Passkeys & remembered access

Manage your existing account security and passkey enrollment.

Verified identity and device support remain required; entering another person's email never grants their access.

Supporting workflow · Operations & account

Notification & digest settings

Manage eligible notification preferences and authorized routes.

Verified destinations and opt-in apply. Saving a setting is not proof of delivery.

Supporting workflow · Data & investigations

Provider connections & sync health

Review connection readiness, mapped sites, synchronization, and source errors.

No provider is connected by the demo; customers retain their own connection state and capabilities.

Supporting workflow · Foundation

Healthy counterfactual API

An implemented backend capability for qualified healthy-production comparisons.

API capability, not a currently implemented standalone customer panel. Access, input evidence, and limitations still apply.

API / operator capability — no standalone web screen.

Supporting workflow · Foundation

Anonymized peer-comparison API

Backend peer comparisons remain separate from Stage 39 portfolio benchmarking.

API capability with cohort/privacy constraints; no standalone web view is currently implemented.

API / operator capability — no standalone web screen.

Supporting workflow · Foundation

Economic-assumption administration

Backend organization settings supply configured rates, costs, and decision thresholds.

Authorized API/configuration workflow. The feature guide does not invent a web editor or change financial assumptions.

API / operator capability — no standalone web screen.

Supporting workflow · Foundation

Fleet audit API

Backend audit capabilities support qualified operational review.

API/operator capability, not a standalone customer dashboard or a completed audit claim.

API / operator capability — no standalone web screen.