Full-Stack Observability for AI Systems.

Logs, metrics, distributed traces, and health checks — unified in one open-source platform. No vendor lock-in. Production-verified.

Explore Features Read the Docs

Part of the BizFirstAi platform · Production-verified March 2026

BizFirst Observe — Grafana

50+

Metrics

Prod sample

Signals

P99

Latency

http_server_request_duration_seconds · P95

INFO PayrollService: run completed acme

WARN Kafka lag > 1000 messages beta

ERR SQL timeout on payroll query acme

INFO Trace exported to Tempo acme

Kafka

Redis

SQL

Loki

Tempo

Grafana

50+

Metrics built-in

Signal types

100%

Dev sampling

Vendor lock-in

What is BizFirst Observe?

Enterprise-grade, unified observability via four signal types. Complete operational visibility into the BizFirstAi platform.

Structured Logs (Grafana Loki)

Serilog structured logs with automatic TenantId, ServerId, RequestId, and TraceId enrichment on every entry. Query with LogQL. Configurable retention from 3 days (development) to 30 days (production).

Time-Series Metrics (Prometheus)

50+ built-in metrics covering every service layer. P50/P95/P99 latency histograms, Kafka consumer lag gauges, and health status gauges. Scrape interval every 15 seconds. Query with PromQL.

Distributed Traces (Grafana Tempo)

Full-chain tracing from HTTP request through service layer to database queries and Kafka. 1% production sampling with always-sample-errors policy. Query with TraceQL. TraceId linked to Loki logs automatically.

Health Checks

GET /health, /health/live, and /health/ready endpoints covering Kafka, Redis, SQL Server, Loki, Tempo, and Grafana. Published automatically as the bizfirst_health_check_status Prometheus gauge. Kubernetes liveness/readiness compatible.

Open-source. No lock-in.

Built on proven open standards. Your data stays in your infrastructure — always.

Built on standards

OpenTelemetry for instrumentation, Prometheus for metrics, Grafana for dashboards. Standard protocols and query languages — not proprietary agents or SPL.

On-prem or cloud

Docker Compose for development and single-server deployment. Kubernetes for production. Runs anywhere you can run a container — no cloud account required.

Your data stays yours

No telemetry sent to a vendor. Loki, Prometheus, and Tempo run inside your own infrastructure. GDPR-ready with configurable TTL retention and no PII in logs.

Three lines of C# to get started.

Full observability — logs, metrics, traces, health checks — with three method calls.

builder.Services.RegisterService_Observability(builder.Configuration);
builder.Host.RegisterSerilog_Observability(builder.Configuration);
app.RegisterApp_Observability();

Explore Features Join Community