Energy Intelligence · Cluster 1

KPIs & CEO report

Reported weekly. As of 2026-09-05. A KPI without a source is not a KPI; a value that is not measured says so.

Revenue€0€ / monthtarget: First paid customer

Pipeline0€ weightedtarget: ≥ 3 qualified opportunities

Qualified customers0counttarget: ≥ 3 (budget owner named, cost quantified)

Pilots0counttarget: 1 signed

Deployments0counttarget: 1

MW / assets connected0MW · assetstarget: First real asset stream

Forecast errornot measuredMAPE / pinballtarget: Beat persistence baseline on customer data

Optimization valuenot measured€ vs baseline scheduletarget: Positive on customer data, CI reported

Latencynot measuredms p95 per tooltarget: Within each tool timeout at p95

Uptimenot measured%target: 99.5 % for read tools

Intervention ratenot measuredhuman overrides / planstarget: Measured before it is optimized

Security findingsnot measuredopentarget: 0 high; public repos scanned for secrets

Research benchmarks0published with CItarget: PINN vs Newton on IEEE 9-bus, published

KPI	Value	Unit	Target	Source	Report section
Revenue	`0`	€ / month	First paid customer	Stripe (energie-teilen) + invoices	`MONEY`
Pipeline	`0`	€ weighted	≥ 3 qualified opportunities	Decision log, customer evidence file	`MONEY`
Qualified customers	`0`	count	≥ 3 (budget owner named, cost quantified)	Customer evidence file	`CUSTOMERS`
Pilots	`0`	count	1 signed	Signed pilot agreement	`CUSTOMERS`
Deployments	`0`	count	1	Kernel deployment registry	`PRODUCT`
MW / assets connected	`0`	MW · assets	First real asset stream	Telemetry ingestion primitive	`PRODUCT`
Forecast error	`not measured`	MAPE / pinball	Beat persistence baseline on customer data	Experiment tracking	`TECHNOLOGY`
Optimization value	`not measured`	€ vs baseline schedule	Positive on customer data, CI reported	optimize_dispatch objective vs baseline_value	`TECHNOLOGY`
Latency	`not measured`	ms p95 per tool	Within each tool timeout at p95	Tool telemetry	`TECHNOLOGY`
Uptime	`not measured`	%	99.5 % for read tools	Observability	`TECHNOLOGY`
Intervention rate	`not measured`	human overrides / plans	Measured before it is optimized	Authorization records	`RISKS`
Security findings	`not measured`	open	0 high; public repos scanned for secrets	Dependency + secret scanning	`RISKS`
Research benchmarks	`0`	published with CI	PINN vs Newton on IEEE 9-bus, published	Research page	`RESEARCH`

CEO report — exactly ten sections, every week

Outcomes, never activity. No commit counts, no repository counts, no lines of code.

Written and maintained by Vincenzo Ceccarelli Grimaldi, Frankfurt am Main.