Industry · Energy & Utilities

Database performance for the grid that can't go dark.

Real-time SQL Server monitoring, AI-assisted query analysis on the on-prem roadmap, and audit-ready telemetry for utility-scale workloads — deployed inside your perimeter, air-gap capable, never routed through a public cloud. Built for the databases that run grid telemetry, AMI, outage management and customer information at utility scale.

NIS2-aligned architectureAir-gap deployablePluggable AI · on-prem or cloud<1% CPU overheadSCADA-side safe
Architected for
NIS2GDPRKVKKAir-gapSCADA-adjacentISO 27001 · Path 2027SOC 2 · Preparation
The problem

The DBA stack at most utilities was built before AMI, before NIS2, before the grid had to answer in milliseconds.

Three pressures that turn utility database observability from an engineering tool into a regulator-level and revenue-level concern.

AMI & telemetry volume

Smart meters now generate billions of reads per day. The databases ingesting, retaining and reporting on them weren't sized for AMI when they were procured. Performance degrades silently — until a regulator asks for a meter-data report that times out.

Risk · operationalAMI scale

Sovereignty & critical infrastructure

SaaS-only monitoring tools cannot run inside critical-infrastructure environments. Shipping grid telemetry to a US public cloud is a non-starter under NIS2, national CI laws and most utility regulators' security baselines.

Risk · regulatoryNIS2

Outage visibility

When the CIS or OMS database slows down, customer-facing outage maps go stale and call centers light up. Legacy tools alert after the SLA is missed — DBAs are the last to know and the first to be blamed.

Risk · customerOutage SLA
Use cases

How utilities and grid operators use Avesys.

Four workloads where sovereign deployment, air-gap capability and utility-scale architecture are non-negotiable.

01 · AMI

AMI database performance at meter-read scale

Smart-meter platforms write hundreds of thousands of reads per minute into SQL Server. DPM tracks ingestion latency, blocking and plan-cache regressions in real time — so a spike in inbound meter traffic doesn't silently corrupt the billing window.

Wait stats · throughputAMI scale
02 · OMS

Outage management system health

When OMS database queries slow down, restoration estimates go wrong and field crews are misdirected. DPM baselines normal OMS query duration and alerts on regression before dispatchers notice the lag.

Baseline · drift · alertBefore dispatch
03 · CIS

Customer information system monitoring

The CIS is where billing, customer service and outage reporting meet. DPM watches the entire SQL Server estate behind your CIS — clusters, AGs, replicas — from one console, with no cloud relay.

Clusters · AGs · replicasOne console
04 · NIS2

NIS2 supply-chain risk by architecture

NIS2 requires demonstrable supply-chain risk control over critical ICT services. Avesys runs entirely on your hardware, with on-prem AI inference on the roadmap — NIS2 supply-chain risk addressed by design, not by attestation.

On-prem AI · no third-party APIBy design
Real-time alerting engine Shipping today

Catch the regression before the outage map goes stale.

DPM watches every monitored instance against thresholds you set and baselines it learns. When AMI ingestion drifts, an OMS query slows or a CIS plan flips, the alert lands in the channel the on-call DBA already lives in — entirely inside your network.

01Threshold & anomaly alerts on instance healthLive
02Routes to the channels DBAs already useLive
03Plan-regression detection on top queriesLive
04Lightweight collector — sub-1% CPU overheadLive
The numbers

What utility CISOs and DBA leads actually measure.

Three numbers a CISO, a DBA lead and a regulator all care about — the cost, the overhead, the supply-chain path. Avesys moves each one in the right direction.

Per instance · per year
€650
Predictable, fleet-scale pricing

One number. No per-database tax, no metric-by-metric upsell. Onboarding the next 30 instances costs what onboarding the last 30 cost — easy to model, easy to defend in a budget review.

Collector overhead
<1%
CPU on monitored instances

The agent is engineered to disappear. Sub-1% CPU on production instances means you can monitor your busiest AMI ingestion database the same way you monitor a BI OLAP node — without becoming the cause of the slowdown.

Third-party cloud dependencies
0
NIS2 by architecture

No outbound API calls. No shared cloud tenancy. The console, the collector and the AI roadmap all run inside the utility perimeter — supply-chain risk addressed by design.

Built for energy & utilities estates

Not a SaaS dashboard with an air-gap promise. Built for the grid.

Three architectural calls that separate database observability built for regulated, critical-infrastructure utilities from a generic monitoring tool.

01 · Sovereignty

Stays inside the utility perimeter, period

No cloud relay. No outbound telemetry. The collector agent sits next to your SQL Server instances and the console runs inside your network. NIS2 supply-chain risk addressed by architecture.

On-prem deployment
Air-gap capable
On-prem AI inference roadmap
No third-party data path
02 · Utility-scale native

Engineered for AMI, CIS and OMS workloads

Designed for SQL Server clusters processing AMI meter reads, OMS dispatch and CIS lookups. Sub-1% overhead even during real outages — the worst time to add observability load.

AMI-scale ingestion Tested
Wait stats first
SCADA-adjacent safe
Deadlock graphs captured
03 · Pricing

One number per instance, every year

€650 per instance per year. Predictable as your estate grows. Pricing scales linearly with AMI rollouts and acquisitions — no per-database tax, no metric-volume billing, no surprise renewal math.

€650 / instance / yr
No per-database tax
No metric-volume billing
Utility-scale licensing
Why this matters now
"A stale outage map is not a database problem. It is a customer-trust problem."
Built by DBAs who have been on-call

Critical infrastructure deserves critical-infrastructure architecture. Not a SaaS sticker.

The founders previously built and operate EST10, an enterprise data management consultancy with five years of experience running mission-critical SQL Server estates for regulated, large-enterprise clients. Avesys is the productization of that experience — built so a utility or grid operator can deploy modern observability without shipping a single byte of grid telemetry to a foreign cloud.

The platform was designed against the bar a serious utility deployment clears — NIS2-aligned, air-gap capable, sub-1% overhead during real outages, and pricing that scales with the estate.

0
Bytes of grid telemetry leaving the network
NIS2 supply-chain risk addressed by architecture — not by attestation.
The Avesys standard for energy & utilities

Six things a database-observability platform must do to belong inside a regulated utility.

Not a feature checklist against other vendors. The bar a serious utility-grade platform clears — and what we built Avesys to deliver from day one.

01 / 06
Sovereign by architectureSovereignty
Deployed entirely inside the customer perimeter. No third-party cloud relay. No telemetry leaves the utility network. NIS2 supply-chain risk addressed by design.
02 / 06
Air-gap capableIsolation
Runs in fully isolated environments. AI inference on the on-prem roadmap. No internet dependency for any monitoring or alerting feature today. Critical for OT-adjacent deployments.
03 / 06
AI-ready, on-prem first
AI-assisted query analysis is on the Avesys on-prem roadmap — built around the constraint that no schema, no query, no plan, and no telemetry leaves your network for the AI to work.
04 / 06
Production-safe at peak load
<1% CPU overhead on monitored instances. Engineered for SQL Server clusters running AMI ingestion, CIS workloads and OMS during real outages.
05 / 06
Audit-ready by default
Every action, threshold change and alert acknowledgment is logged with operator attribution and an exportable audit trail. Regulator questions answer themselves.
06 / 06
Affordable at utility scale
A utility with 100+ SQL Server instances should not spend six figures on monitoring. Pricing scales linearly at €650 per instance per year.
Pricing

One line item. Predictable as the grid grows.

Per-instance pricing, every year. Everything DPM ships today — wait stats, blocking, queries, fleet view, alerting, the lightweight collector — included.

Avesys DPM · per SQL Server instance
€650 / yr
All features included · no add-on SKUs

One annual fee per monitored instance. Whether the instance powers AMI ingestion, OMS dispatch, CIS or BI — same number, same product, same agreement.

Real-time instance monitoring
Wait statistics analysis
Blocking & deadlock detection
Query performance analysis
Threshold & anomaly alerting
Multi-instance fleet view
Lightweight on-prem collector
No outbound telemetry
Sample · 80 instances
€52,000 / yr
Regional utility

Roughly the SQL Server footprint of a regional electricity or gas utility — AMI, CIS, OMS and BI. Full DPM feature set on every instance.

Sample · 250 instances
€162,500 / yr
National grid operator

Same per-instance number — the only thing that changes is how many instances you're monitoring. Procurement-ready: the line item scales linearly with the estate.

Calls to action

Pick your starting point — we'll meet you there.

Three low-friction ways to start a conversation. Whichever you choose, you'll land on a 15-minute call with someone who has actually been on-call for a utility SQL Server estate.