Pipelines as code
Define ingestion and transformation steps in versioned configuration. Review changes like any other code.
Self-hosted data platform
Your self-hosted data platform.
Your servers.
No subscription meter.
Stop renting your data platform. Own it.
Prooflix lets your team build pipelines that run Spark jobs through a medallion architecture, installed on infrastructure you already own. No per-seat licences, no compute markup, and no data leaving your network.
- Runs on Kubernetes, VMs or bare metal
- Installs with any cloud or SaaS provider, or fully on-premise
- Open table formats, no lock-in
- Single installation for the whole team
Prooflix is at an early stage. The capabilities described on this page are the project's direction and are still being built. There is no company behind it yet.
Runs where your data lives
Install inside your own network or on any cloud provider you already use. Nothing is uploaded to a vendor-owned platform.
Pay for hardware, not usage
No per-seat fees and no compute credits. The cost is the servers you already run.
Standard tools, open formats
Spark, SQL and open table formats. Your data stays readable by any engine.
Why Prooflix
Everything a hosted data platform gives you, without renting it.
Managed SaaS platforms are convenient until the invoice grows with every user, every query and every gigabyte of egress. Prooflix brings the same workflow in-house.
Hosted SaaS platform
- Monthly bill scales with seats, compute and storage
- Data is copied into the vendor's cloud account
- Egress and cross-region traffic are billed separately
- Proprietary runtime makes migration costly
- Tied to the hosting provider the vendor chose
- Compliance reviews for every new workspace
Prooflix, self-hosted
- One installation, flat cost, unlimited users
- Data never leaves your network or your controls
- No egress charges between your own systems
- Plain Spark and open formats, portable by design
- Installation complies with any cloud or SaaS provider, or on-premise
- Fits the security posture you already have
Platform
Built for pipelines, layered data and Spark processing.
Medallion architecture built in
Bronze, Silver and Gold layers are first-class. Promote data between layers with validation at each step.
Spark inside every pipeline
Each pipeline can run Spark jobs on your cluster. Jobs queue, scale and retry as part of the pipeline run.
Scheduling and dependencies
Run on a cron schedule or on upstream completion. Backfill ranges and re-run failed steps only.
Observability
Logs, run history, row counts and durations for every task, in one place, kept on your own storage.
Open formats and governance
Tables stay in open formats with schema history and access controls that plug into your directory.
How it works
From empty servers to a running lakehouse in four steps.
-
1
Install
Deploy Prooflix to your Kubernetes cluster or a set of VMs with a single command. The installation complies with every major cloud and SaaS provider, and runs just as well fully on-premise. Storage can be local disks or any S3-compatible object store.
-
2
Connect sources
Point Prooflix at your databases, message queues, file drops and internal APIs using built-in connectors.
-
3
Define pipelines
Describe how data lands in Bronze, how it is cleaned into Silver and how it is shaped into Gold tables for analytics.
-
4
Run and watch
Pipelines run their Spark jobs on a schedule or on demand. Track every run from the dashboard.
Cost model
One installation. Zero usage meter.
Hosted platforms charge for every unit of compute, every active user and every byte moved. Those costs rise with adoption, so success gets expensive. Prooflix runs on hardware you already budget for, so adding a team or a pipeline does not add a line to the invoice.
- No per-user or per-workspace licences
- No compute credits or runtime surcharge
- No egress fees inside your own network
- Predictable spend that finance can plan around
FAQ
Questions teams ask before leaving a hosted data platform.
What is Prooflix?
Prooflix is a self-hosted data platform for building data pipelines and running a medallion architecture. You install it on your own servers, Kubernetes cluster or cloud account instead of renting a hosted SaaS data platform.
Can I run a medallion architecture on-premise?
Yes. Prooflix implements the Bronze, Silver and Gold medallion layers out of the box. Raw data lands in Bronze, is cleaned and validated into Silver, and is curated into Gold tables for analytics, all inside your own infrastructure.
How do Spark pipelines run in Prooflix?
Each pipeline can run Spark jobs on your cluster. Jobs are scheduled, queued and retried as part of the pipeline run, so there is no separate job server to manage and no compute credits to buy.
Does Prooflix work on AWS, Azure, Google Cloud or fully on-premise?
The installation complies with any cloud or SaaS provider and also runs fully on-premise on VMs or bare metal. Storage can be local disks or any S3-compatible object store.
How does Prooflix avoid SaaS subscription costs?
There is one installation with no per-user licences, no per-workspace fees and no compute markup. You pay for the hardware you already run, so adding users or pipelines does not increase the bill.
Is there vendor lock-in?
No. Prooflix uses plain Spark and open table formats, so your data stays readable by any engine and you can migrate at any time.
Is Prooflix a company?
Not yet. Prooflix is an early-stage project maintained by a small group of contributors. There is no registered company, no sales team and no paid plan. The features on this page describe where the project is heading, and some are still in development.
Contact
Leave your details and we will get back to you.
Tell us about your data platform plans, ask a question, or offer to contribute. Messages go straight to the Prooflix team inbox and we reply by email.
Want to follow or help shape Prooflix?
Prooflix is an early-stage project, not a company. Try the current build, share feedback, or contribute to where it goes next.