z4j Stable
Z4J
The control plane for Python task queues.
One dashboard for Celery, RQ, Dramatiq, Huey, arq, and taskiq. History, controls, and schedules, without touching your task code.
PyPV // Python Platform Ventures
Python-native control planes for infrastructure, automation, and AI agents.
PyPV is an independent developer studio building open-source, self-hosted systems on Python and FastAPI for background jobs, multi-vendor networks, IP KVM and server fleets, and the human decisions AI agents can't make alone.
- python
- fastapi
- sqlalchemy
- postgres
- react
What PyPV builds
See all projects
Independent systems, shared operational DNA.
-
- freesdn Early access
FreeSDN
One platform. Every vendor.
Self-hosted central command for multi-vendor infrastructure. Observe and act from one web UI and one REST API.
- ony Public beta
Ony.ai
The on-call layer for AI coding agents.
When your agent needs you, Ony.ai calls your phone. Answer from anywhere, and the agent keeps moving.
- qkvm Public beta
QKVM
Every machine on one wall.
A self-hosted screen wall for IP KVMs and servers: the live screen and state of every machine you run, on one page.
Capabilities
The operational layer, end to end.
From background jobs and network state to bare-metal consoles, and the human decisions autonomous software can't make alone.
-
Background jobs
Visibility and control for the work that runs behind your product.
- Task history
- Worker views
- Safe retries
- Schedulers
- Failure grouping
-
Networks & infrastructure
One control plane for multi-vendor, self-hosted infrastructure.
- Multi-vendor adapters
- Staged writes
- Multi-tenant RBAC
- Site connectivity
- Zero-touch provisioning
-
Agents & human-in-the-loop
The on-call layer between coding agents and the people who decide.
- Coding agents
- Phone handoff
- Risk classification
- Signed verdicts
- Audit trails
-
Hardware & remote access
Every KVM, BMC, and server you run, on one live wall.
- Screen walls
- IP KVMs
- Redfish BMCs
- Bulk power actions
- Wake-on-LAN
How PyPV builds
Principles, not promises.
A consistent way of building that holds across every project.
- 01
Operational by default
Every system ships with the visibility operators need: state, history, audit trails, and live status. If you can't see it, you can't run it.
- 02
Self-hosted, open source
Every project runs on your own infrastructure and publishes its source under the AGPL. Python and FastAPI at the core, no mandatory cloud, no vendor lock-in.
- 03
Human-in-the-loop
Automation handles the repeatable work and hands consequential decisions back to people. Agents propose, humans stay in control.
- 04
Built to last
Typed, tested, and deliberately boring where it counts. The goal is infrastructure that runs quietly in production for years, not a demo.
Technical focus
The layer PyPV works in.
A focused stack for control planes, workers, and devices: the operational systems that keep automation accountable.
- Python
- FastAPI
- SQLAlchemy
- PostgreSQL
- Valkey
- Celery
- React
- TypeScript
- WebSockets
- Docker
- Prometheus
- Redfish
- KVM-over-IP
- Telephony
- Network automation
- AI coding agents
- Control planes
Philosophy
Built for the parts of software that cannot just be a prompt.
PyPV focuses on the operational layer around AI and automation: queues and workers, network state, machine consoles, approvals, and observability. The goal is autonomous systems reliable enough to run in the real world, with people still in charge of the decisions that matter.
Building something Python-native?
Reach out about collaboration, open-source infrastructure, AI-agent workflows, or a technical partnership. Every message gets read.