Engineering / projects
Some projects are already running; others are being built or are still ideas. This registry records where each one stands.
These services are in operation. The catalog describes what they do; addresses and operational data remain private.
GPS tracking platform and device management.
Infrastructure visibility and monitoring.
Virtual infrastructure and cluster administration.
Patch management and security visibility.
Network source of truth, IPAM, and DCIM.
Distributed job scheduling and execution.
Service availability monitoring.
Hosted web servers for client projects.
PBX voice infrastructure based on the SIP protocol.
An intrusion detection system independent of major alarm vendors and their cloud services.
Under development using Home Assistant for automation, Alarmo for alarm logic, Zigbee for wireless sensors, and Tailscale for private connectivity. The components are still being integrated; the system is not complete.
A video recorder with locally managed storage and access to footage.
A DVR under development to record, store, and review footage on directly managed infrastructure instead of a vendor cloud. Its final capabilities are still being worked out.
Documentation stored as linked information rather than separate text files.
The idea is to store facts, decisions, dependencies, and procedures as linked entries. Pages would draw from those entries, allowing a fact to be updated once wherever it appears.
Search that takes both the question and its context into account.
A question can mean something different during an incident than during a change review or onboarding. Search would use that context to find relevant information, rather than relying on keywords alone.
Network configurations as typed objects, not text files.
Model interfaces, VLANs, policies, and routes as related objects. Generate device configuration from that model and validate changes before applying them.
Infrastructure configuration kept in version control and reviewed before use.
Describe the intended state of compute, networking, and services in code. Review changes before applying them, provide a rollback path, and report differences between the running system and its declared state.
Scheduled restore tests to check that backups can be recovered.
Restore selected datasets into an isolated environment on a schedule. Record the result and compare recovery times with the target.
Publish service availability without exposing internal addresses.
A public status page showing availability without disclosing service endpoints or internal topology. Access to the services would remain restricted.
Building means work is underway. Considering and Sketch are ideas, not deployed systems. None has a promised delivery date.