Use Cases¶
Anosys is a universal observability platform. Below are practical use cases showing how teams use it across different domains — from website performance to AI agent workflows and infrastructure monitoring.
Website Traffic & Performance Monitoring¶
Track page views, user sessions, scroll engagement, and server-side latency for any custom website — without heavyweight analytics SDKs.
| What You Can Track | How |
|---|---|
| Page views, referrers, sessions | JavaScript tag or standalone image pixel |
| Scroll depth & engagement time | Automatic via JS tracker |
| Server response time & errors | REST API from your backend |
| Custom business events | String, numeric, and boolean fields piggybacked on the URL |
With Anosys, you get client-side and server-side data in one platform — correlated automatically. Add ML-based anomaly detection and Slack/email alerting on top, and you have a full-stack web monitoring setup in minutes.
→ Read the full tutorial: How to Track Website Traffic and Performance with Anosys
Agentic AI Observability¶
AI agents are non-deterministic, multi-step, and expensive. A single coding session can consume thousands of tokens with no visibility into what went right — or wrong. Anosys provides end-to-end tracing for agentic workflows.
| What You Can Track | How |
|---|---|
| Model invocations & tool calls | Native SDKs or OpenTelemetry |
| Token usage & cost attribution | Automatic per-request tracking |
| Latency breakdowns | Per-call and end-to-end timing |
| Error classification | Structured logs with auto-grouping |
| Model parameter tracking | Temperature, top-p, model version |
Supported frameworks include OpenAI Agents, Anthropic/Claude Code, and any LLM provider via OpenTelemetry or the REST API.
→ Guides: OpenAI Agents · OpenAI ChatKit Apps · Anthropic Agents · Custom LLM Integrations
Full-Stack AI Observability¶
Most AI monitoring tools only see the model layer. Most APM tools only see infrastructure. Anosys bridges the gap with true end-to-end AI observability — unified visibility across user behavior, AI agents, application services, and infrastructure in a single platform.
| What Sets Anosys Apart | Why It Matters |
|---|---|
| Cross-layer correlation | Trace a user complaint through the UI, the agent's reasoning, the backend API, and the underlying infrastructure — in one view |
| User behavior tracking | Monitor how real users interact with AI features — session flows, engagement, abandonment — using a lightweight JS tag or image pixel |
| Automated anomaly detection | ML-based baselines detect silent failures, cost spikes, and quality regressions across every layer — no manual thresholds required |
| Root cause analysis | Go from "something broke" to "here's why" in minutes with causal paths that span agents, models, and infrastructure |
| Custom tracking fields | Send arbitrary string, numeric, and boolean fields to capture any business-specific signal — all automatically indexed and queryable |
Unlike point solutions such as Arize (model-only), Dynatrace (infra-only), or Google Analytics (traffic-only), Anosys gives you the complete picture so you can see what other vendors can't.
→ Read the full analysis: What Is AI Observability — And Why Current Tools Are Failing You
Network & Infrastructure Monitoring¶
Monitor routers, switches, firewalls, servers, and containers at scale. Anosys accepts metrics from any device that can make an HTTP call or export OpenTelemetry signals.
| What You Can Track | How |
|---|---|
| Interface utilization, packet loss, error rates | REST API or OTEL Collector |
| CPU, memory, disk I/O, network throughput | OpenTelemetry or custom agents |
| Container health & pod restarts | Kubernetes OTEL integration |
| Request rates, error rates, p95 latency | Application-level instrumentation |
Deploy an OpenTelemetry Collector as a central aggregation point, or use lightweight REST API calls from custom device agents.
→ Guide: Network & Infrastructure Observability
IoT & Edge Device Monitoring¶
Collect telemetry from sensors, edge gateways, and device fleets at massive scale. Track heartbeats, firmware versions, connectivity status, and environmental readings.
| What You Can Track | How |
|---|---|
| Temperature, humidity, voltage readings | REST API (HTTP GET/POST) |
| Device heartbeats & connectivity | Periodic API calls or image pixels |
| Fleet firmware versions | Custom string fields |
| Message queue depth & processing latency | OpenTelemetry or REST API |
Anosys handles millions of data points per second without sampling or data loss.
Custom Business KPIs¶
Track any business-specific metric that matters to your team. Anosys accepts arbitrary string, numeric, and boolean fields — all automatically indexed and queryable.
Examples:
- E-commerce — cart abandonment rate, checkout latency, payment success rate
- SaaS — feature adoption, API error budgets, customer health scores
- Media — content engagement, ad fill rates, video completion rates
- Finance — transaction anomalies, SLA compliance, fraud pattern detection
Use custom pipelines to transform, enrich, and route your business data in real time or on a batch schedule.
Getting Started¶
All use cases start the same way:
- Sign up at console.anosys.ai
- Create a pixel — choose the integration type that fits your use case
- Send data — follow the relevant guide above or the Data Ingestion Options
- Explore — dashboards populate within seconds
Last updated: March 8, 2026