LynxOps
An AI-powered enterprise observability and incident management platform that monitors SAP NetWeaver / HANA / SQL Server and OS infrastructure in real time.
Overview
Unifies the scattered monitoring tools of SAP Basis and infrastructure teams into a single pane; turns ABAP dumps (ST22), HANA memory/delta merge, business processes, and OS metrics from reactive to proactive through correlation and anomaly detection. Target environment: on-premise / air-gapped — your data stays with you.
SAP-specific depth (ST22 dump AI analysis, HANA delta merge, work-process heatmap) is rare in general APM tools; data sovereignty with air-gapped + local Ollama; agent evolution from Python to Go; enterprise HA and security architecture.
In production. 30+ API routers, 40+ Alembic migrations, 50+ pytest test files, GitHub Actions CI, air-gap and Helm support.
Highlights
H01–H14 + live topology: admin view, SLA & analytics, infrastructure cockpit, DB monitor (HANA/SQL Server), SAP app/user/audit, automation.
An AI pipeline that analyzes SolMan/CloudALM EarlyWatch reports and ST22 dumps with Claude + local Ollama.
Z-Score-based correlated detection; ST22→STMS incident correlation and maintenance windows.
SAP rules, automatic action triggers, approval flow.
Bulk ack/close, audible alerts, PWA + web push, mobile actions, on-call schedule.
ServiceNow, PagerDuty, OpsGenie, Jira.
5-component global health score, scheduled PDF/Excel reports.
<20MB with offline buffer; migration from Python agents to a Go agent — SAP/HANA/SQL/OS plugins.
SAML SSO, action-level RBAC, MultiFernet encryption + JWT key rotation, immutable audit log, multi-tenancy.
Tech Stack
| Backend | FastAPI · SQLAlchemy 2 · TimescaleDB (PG16) · Redis Sentinel (HA) |
| Task queue | Celery / RedBeat |
| Frontend | Next.js 14 · React · TanStack · Zustand · Recharts |
| AI | Anthropic Claude + Ollama |
| Agent | Go edge agent (<20MB, offline buffer) |
| Infrastructure | Docker · Kubernetes (Helm) · Nginx |