>_
Fast-Flow
01 deploy 02 product 03 anatomy 04 config 05 compare manifesto
GitHub {{ stars }}
{{ version }} · open source · GPL-3.0 · Docker + Kubernetes

The orchestrator for people who just have scripts.py

No DAGs. No image builds. No YAML. A pipeline is a folder with a main.py — push it to Git and it runs in its own disposable sandbox, dependencies resolved in milliseconds.

Star on GitHub{{ stars }}
38ms
cached deps, linked
1
process to operate
0
images ever built
← pick a scenario
{{ termTitle }}
{{ l.text }}
01 / deployment

The ten-minute deploy, deleted.

You changed one line in requirements.txt. Here's what happens next — in a classic image-build pipeline versus Fast-Flow. Press deploy.

simulated · 1s ≈ 45s of real time
Image-build pipeline {{ tradClock }}
{{ st.icon }} {{ st.label }} {{ st.dur }}
Fast-Flow {{ ffClock }}
{{ st.icon }} {{ st.label }} {{ st.dur }}
{{ raceVerdict }}
02 / product

A control room, not a config maze.

live mock — click the sidebar
>_
FastFlow
Overview
Orchestration
Security
All systems operational
{{ demoCrumb }} Search or jump to… ⌘K
{{ k.label }}{{ k.delta }}
{{ k.value }}
Run activitylast 26 weeks · 412,338 runs
{{ p.name }}
{{ p.cron }}next {{ p.next }}
PipelineStatusSuccess rateScheduleLast run
{{ p.name }} {{ p.status }} {{ p.rate }} {{ p.cron }} {{ p.last }}
run_8f3a2c RUNNING daily_etl · py3.12 · 0.5 CPU / 512m
AllstdoutstderrFollow
{{ l.n }}{{ l.text }}
● streaming via SSE{{ logCount }} lines
Encrypted at rest with Fernet · never shown in logs ⚿ encryption key loaded
KeyValueScopeUpdated
{{ sr.key }} •••••••••••• {{ sr.scope }} {{ sr.updated }}
Aggregated from all requirements.txt · CVE check nightly 2 advisories
PackageVersionUsed bySecurity
{{ dr.pkg }} {{ dr.ver }} {{ dr.used }} {{ dr.cve }}
ScheduleCronEnvironmentNext runEnabled
{{ sr.name }} {{ sr.cron }} {{ sr.env }} {{ sr.next }}
03 / anatomy

Six decisions that make it fast.

{{ featFile }}{{ featTag }}
{{ featCode }}

{{ featBody }}

04 / configuration

The entire config surface fits on this page.

Resource limits, retries, timeout, Python version — one JSON file next to your code, versioned in Git. Try it:

CPU limit
Memory limit
Python version
Retries{{ retries }}
Timeout{{ timeoutLabel }}
Schedule
Env variable · LOG_LEVEL
Custom env variable
=
Merged into default_env — secrets belong in the Secrets UI, not here.
Soft limits are monitored and surfaced as warnings in the UI; hard limits are enforced by cgroups (Docker) or resources.limits (K8s). Exceed memory → OOM kill, exit 137, visible in the run detail.
pipelines/daily_etl/pipeline.json 
{{ cfgJson }}
05 / positioning

"A handful of Python scripts, and Airflow is overkill."

If that sentence is you, this is your tool. If not — we'll tell you what to use instead. Pick a comparison:

Fast-Flow{{ compName }}
{{ r.label }} {{ r.ff }} {{ r.other }}
When you should use {{ compName }} instead

{{ compCaveat }}

06 / quickstart

Running before your coffee is.

Clone, set one encryption key, compose up. Then open localhost:8000 and push your first pipeline from the template repo.

{{ qsFile }} 
{{ qsCode }}
>_
Fast-Flow Orchestrator · GPL-3.0 · {{ version }}
GitHub Docs Manifesto Releases