OIS - Observability Intelligence System

01 / Ingest

Universal Data Ingestion

Capture every signal from your infrastructure. OIS ingests metrics, logs, and traces through OpenTelemetry with automatic schema detection and intelligent routing.

OTLP native with gRPC and HTTP support

Multi-tenant isolation at the edge

Per-tenant quotas and rate limiting

Sub-second ingestion latency

Custom JavaScript ingestors with hot-reload

02 / Dashboards

Visualize Everything

Build stunning dashboards with drag-and-drop simplicity. From real-time metrics to historical trends, see your entire system at a glance with customizable widgets and layouts.

03 / AI Terminal

Build with Intelligence

Describe what you need in plain English. The AI terminal understands your intent and generates dashboards, functions, queries, and custom views instantly.

OIS AI Terminal

ois> Create a dashboard showing CPU and memory usage across all hosts with alerts for when usage exceeds 80%

Analyzing request... Creating dashboard with 2 metric panels and 2 alert rules.

✓ Created dashboard: "Infrastructure Health"
✓ Added panel: CPU Usage (gauge + time series)
✓ Added panel: Memory Usage (gauge + time series)
✓ Created alert: cpu_high_usage (threshold: 80%)
✓ Created alert: memory_high_usage (threshold: 80%)

ois> Write a function that sends a Slack notification when error rate spikes

Generating serverless function with Slack webhook integration...

✓ Created function: error-rate-alerter.js
✓ Configured trigger: error_rate > 5% for 5 minutes
✓ Ready to deploy

ois> Investigate the payment-service alerts firing for the last 30 minutes

Analyzing 3 active alerts... Querying metrics, logs, and traces for payment-service.

⚠ ALERTS: error_rate_high (23%), latency_p99 (2.4s), connection_pool_exhausted
■ METRICS: DB connection wait time spiked 340% at 14:23 UTC
▸ LOGS: 1,847 "connection timeout" errors from postgres-primary
◈ TRACES: Upstream bottleneck at inventory-service → db-pool

Root cause: inventory-service deploy at 14:21 UTC introduced N+1 query regression.

ois> Fix it

Rolling back inventory-service to v2.3.1...

✓ Rollback complete: inventory-service v2.4.0 → v2.3.1
✓ Connection pool recovering: 23/50 → 48/50 available
✓ Error rate: 23% → 0.4%
✓ All 3 alerts resolved

Incident resolved in 4m 12s. Post-mortem report generated.

ois> _

□

Dashboards

Generate complete dashboards from descriptions

❮/❯

Functions

Write and deploy serverless automation

∑

Queries

Translate questions into SQL instantly

⚙

Custom Views

Build interactive HTML visualizations

⚛

Agentic Operations

A frontier AI agent with tool execution—autonomously explore data, run queries, and resolve incidents

04 / Code Blocks

Query with Power. Visualize at Scale.

Write sophisticated queries and render stunning visualizations directly in the browser. Code Blocks combine full SQL power with hardware-accelerated graphics—from simple charts to complex 3D particle systems, all at 60fps.

Full SQL support with ClickHouse backend

Hardware-accelerated rendering (WebGL/Canvas)

Intelligent autocomplete and syntax highlighting

Save, share, and schedule queries

05 / Functions

Serverless Automation

Deploy JavaScript functions and AI-powered tasks. Schedule a nightly anomaly report, auto-triage alerts with an AI agent, or trigger webhooks—without managing infrastructure.

06 / Files & Custom Views

Generate Custom Views

Create and manage custom HTML views, reports, and visualizations. Upload assets, generate dynamic content, and serve interactive dashboards tailored to your needs.

07 / RBAC

Enterprise Security

Fine-grained role-based access control keeps your data secure. Define custom roles, assign granular permissions, and maintain complete audit trails across your organization.

JSON-based policy engine for fine-grained control

SSO/OIDC integration

API key management with expiration controls

Complete audit logging

Observability Intelligence System