Data Pipelines & IPAM Integration · NetBox Consultant

Infrastructure Documentation as Code

If your infrastructure automation still relies on spreadsheets as the source of truth, you’re building on sand.

Every automation initiative depends on accurate infrastructure data. When the source of truth is stale or scattered across spreadsheets, every playbook needs manual verification and every deployment carries risk. IPAM/DCIM is the foundation for infrastructure as code, and treating it as a documentation project is how it ends up stale.

What I do is make NetBox the intended state rather than a documentation tool: automated imports from discovery sources, validation pipelines that flag drift, and change workflows that update NetBox first and provision second.

From Discovery to Source of Truth

NetBox itself comes first, meaning deployment, data model design, and the initial data import, plus integration with existing systems over the REST API and whatever custom scripts and reports the workflows call for.

Where dcTrack is also in play, the two need to agree. I build DCIM synchronization between NetBox and dcTrack that keeps asset, power, and location data consistent across systems that serve different stakeholders.

Feeding those systems means automated data collection from network devices such as Juniper and Cisco, from hypervisors including VMware vSphere and Proxmox, and from other infrastructure sources, with reconciliation logic that handles the messy reality of production environments.

The ETL pipelines themselves use Prefect for orchestration, Pandas for transformation, and DuckDB for analytical queries over infrastructure data, running on a schedule so the source of truth stays current. Taken together they consolidate infrastructure information from network devices, hypervisors, and virtualization platforms into an authoritative inventory that gives teams a single, current view to work from.

The last piece is drift detection: pipelines that compare intended state in NetBox, dcTrack, or another DCIM solution against actual state in network, compute, and storage, alert on discrepancies, and remediate automatically where that’s appropriate.

What Projects Look Like

• Greenfield NetBox deployment with data migration from spreadsheets
• Integration pipelines connecting NetBox/dcTrack to Ansible for inventory
• dcTrack/NetBox synchronization for datacenter operations
• Network device discovery and import automation
• Custom dashboards for infrastructure visibility (Streamlit, Dash/Plotly, Grafana)