Anomaly detection
Learning models that learn your normal curves and alert on real deviations — not on noise.
AI-augmented observability·EU sovereignty 🇪🇺
See it all. Understand it all. Fix it before it breaks.
lux-aiops unifies metrics, traces and logs in a single OpenTelemetry-native platform. Our AI detects anomalies, finds the root cause and triggers remediation — automatically. No proprietary SDKs, no lock-in.
Hosted in the EUSOC 2 · ISO 27001 — in progressNo credit card · 14 days
app.lux-aiops.com/observe
LiveAuto-detected anomalies
1 284Real-time
Median MTTR
4m 12s-63%
p99 latency
118 msstable
Request throughput — production6h
api-gatewaycheckout-svcanomaly
AIOps Copilot
Latency spike on checkout-svc correlated with deployment #4f1a. Likely cause: DB pool saturation.
312
Services
99.99%
Uptime
Trusted by 600+ engineering teams across Europe
NebulaPayHelios CloudBanque du Grand-DuchéOrbit LogisticsKairos HealthVolt MobilitySaphir TelecomAtlas RetailNebulaPayHelios CloudBanque du Grand-DuchéOrbit LogisticsKairos HealthVolt MobilitySaphir TelecomAtlas Retail
One platform, not ten tools
Everything you need to operate with confidence
Metrics, traces, logs, security and AI — together, to cut your resolution time and your mental load.
Root-cause analysis
Automatic correlation of metrics, traces and logs to pinpoint the cause of an incident in seconds.
Automatic remediation
Trigger runbooks, rollbacks or scalings as soon as a threshold is crossed. Fixed before on-call wakes up.
Service map
A real-time topology of your dependencies, generated from distributed traces. No manual setup.
Logs without sampling
Ingest, index and full-text search 100% of your logs, with flexible retention and controlled cost.
Synthetic monitoring
Simulated user journeys from 12 regions to catch outages before your customers do.
And also:Deployment trackingKubernetes auto-discoveryThreat detection
The platform
One unified console for your whole stack
No more jumping between five tools during an incident. All your context in one place.
Dashboards that answer “why?”
Visualize the health of every service, database and queue. Zoom from the big picture down to the exact trace slowing your p99.
- End-to-end distributed tracing
- Continuous code profiling in production
- Over 750 ready-to-use integrations
Latency per endpoint● healthy
p50
41 ms
p95
96 ms
p99
118 ms
OpenTelemetry-native architecture
Built on open standards
European enterprises in banking, manufacturing and the public sector will never be 100% public cloud. lux-aiops treats on-premise as a first-class citizen — with OpenTelemetry® Collector agents and open protocols end to end, not a proprietary agent you'll have to rip out.
The cost killer
Edge-filtered telemetry
Every node runs an OpenTelemetry Collector agent: OTTL filter and transform processors drop noise at the source, and policy-based tail sampling keeps the traces that matter — errors, latency outliers, rare paths — before a single byte leaves the host.
75–90%
less ingested trace volume — the range industry deployments report with edge tail sampling
NIS2-ready
Zero-egress by design
Agents ship OTLP over your internal network to gateway collectors — the documented OpenTelemetry agent-to-gateway pattern for restricted-egress and air-gapped environments. Filtering, correlation and AI inference run inside your perimeter; the gateway is the single, auditable egress point — or none at all.
Zero-friction
Fleet management over OpAMP
Our Discovery engine maps unmanaged nodes and provisions collectors; from then on the fleet is supervised via OpAMP, the emerging OpenTelemetry agent-management standard — centralized configuration rollout, live health reporting, last-known-good resilience.
Already on OpenTelemetry?
Send us OTLP. That's the whole integration.
Point your existing OpenTelemetry Collector — or any OTel SDK — at our EU-hosted endpoint (gRPC 4317 / HTTP 4318). Your semantic conventions are preserved end to end, and you can export everything back out via OTLP whenever you choose. No proprietary SDKs, no lock-in.
OpenTelemetry® and the OpenTelemetry logo are registered trademarks of The Linux Foundation.
# otel-collector.yaml
exporters:
otlphttp:
endpoint: https://ingest.eu-lux-1.lux-aiops.com
headers:
authorization: Bearer ${env:LUX_AIOPS_TOKEN}
service:
pipelines:
traces:
exporters: [otlphttp]# …or a single environment variable
OTEL_EXPORTER_OTLP_ENDPOINT=https://ingest.eu-lux-1.lux-aiops.comStrategic positioning
| Capability |  Legacy (Datadog) | New SaaS (Tsuga) | lux-aiops |
|---|---|---|---|
| Architecture | Vendor SaaS cloud | Client cloud (VPC) | Hybrid: sovereign cloud + on-prem edge |
| Agent model | Proprietary agent, high-volume streaming | Agent-based (cloud-focused) | OpenTelemetry Collector distro — swap it anytime |
| Cost driver | Punitive per-GB ingest tax | Compute-based | Predictable flat fee (zero external egress) |
| Scope | Observability & security | AI-native observability | Discovery + AIOps + on-prem + ITSM |
| Compliance | Standard SOC 2 / GDPR | Data localization | Native, heavy NIS2 compliance |
The pitch for CIOs
“You're currently paying premium cloud ingestion rates for on-premise data that doesn't need to leave your building. lux-aiops runs standard OpenTelemetry Collectors that filter your telemetry locally, connect directly to your ITSM for automated resolution, and keep your data firmly within your sovereign boundaries — and because ingestion is pure OTLP, you can point your existing OTel setup at us today and take your data back out the day you choose.”
−63%
lower mean time to resolution (MTTR)
12 PB
of telemetry ingested every day
99.99%
uptime guaranteed by SLA
< 30s
to detect a critical anomaly
Integrations
Connected to your whole stack in minutes
Over 750 native integrations. Plug in your sources, auto-discovery does the rest — no agent to configure by hand.
European sovereignty
Your data stays with you, in Europe
lux-aiops is designed and operated in Luxembourg. Where US platforms expose your data to foreign jurisdictions, we guarantee full sovereignty — with no compromise on performance.
Data hosted in Luxembourg
Your telemetry stays in Tier IV datacenters at the heart of the EU. No data ever leaves European soil.
Compliant by design
GDPR & NIS2 by design; SOC 2 Type II, ISO 27001 and DORA alignment in progress. End-to-end encryption, in transit and at rest.
No extra-EU dependency
Infrastructure and support operated in Europe, shielded from extraterritorial access laws.
🇱🇺
Data region
eu-lux-1
Latency < 8 ms from Paris, Frankfurt, Brussels and Amsterdam.
GDPR
Compliant
DORA
In progress
NIS2
Covered
ISO 27001
In progress
Trusted by teams
The teams that sleep better at night
“We cut our MTTR by three in a quarter. The copilot pinpoints the root cause before we've even finished reading the alert.”
Camille Roussel
VP Engineering, NebulaPay
“Finally an observability platform hosted in Europe. Our CISO signed off in two weeks what used to take months with a US vendor.”
Tobias Wagner
Head of Platform, Helios Cloud
“We replaced three tools with lux-aiops. Fewer licenses, less context-switching, and a finally predictable telemetry bill.”
Amina Cherif
SRE Lead, Volt Mobility
Founders
Built by operators, for operators
A team that has run production at European scale.
Wilfried Lucas
Co-founder
Manager of Production Tools · POST Luxembourg
Pricing
Pricing that is clear and predictable
Pay as you go, with no surprises on your telemetry bill. Change plans at any time.
MonthlyAnnual −20%
Starter
For small teams getting started.
Free
Start for free- 5 hosts included
- Metrics & dashboards
- 3-day log retention
- Basic alerting
- Community & docs
Most popular
Pro
For teams that are scaling up.
€23/host/month
Start the trial- Per host / month
- APM & distributed tracing
- AI copilot & root cause
- 15-day log retention
- Synthetic monitoring
- Priority support
Enterprise
Sovereignty, SLA and advanced compliance.
Custom
Contact the team- Negotiated volume & price
- Dedicated eu-lux-1 region
- SSO/SAML & advanced RBAC
- 99.99% SLA & 24/7 support
- DPA, DORA & dedicated audit
FAQ
Questions, answered
In Tier IV datacenters in Luxembourg (EU). No telemetry ever leaves European soil, and the infrastructure is operated entirely in the EU.
Yes. Ingestion is pure OTLP (gRPC 4317 / HTTP 4318) — point your existing OpenTelemetry Collector or SDKs at our EU endpoint and you're done. Our edge agent is a hardened OpenTelemetry Collector distribution, your semantic conventions are preserved end to end, and you can export your data back out via OTLP at any time. No proprietary SDKs, no lock-in.
Our agents filter telemetry at the edge and forward only what matters, so you avoid punitive per-GB ingest and external egress fees — a predictable flat fee instead of an exploding bill.
Yes. Agents run in fully air-gapped environments and route data through internal secure gateways to your self-hosted or sovereign control plane.
Minutes for cloud sources via auto-discovery. A typical POC is validated in two weeks — no large consulting project required.
GDPR and NIS2 by design — EU-hosted, your data never leaves the Union, with end-to-end encryption in transit and at rest. SOC 2 Type II, ISO 27001 and DORA alignment are on our roadmap, not yet certified.
750+ native integrations (AWS, Azure, GCP, Kubernetes, Kafka, PostgreSQL, OpenTelemetry…) plus your ITSM for automated resolution.
Take back control of your production
Get started in under 5 minutes. No credit card. Your data stays in Europe, from the very first metric.
14-day trial · No complex install · Uninstall in one click