Managed Workflows
Describe the job. Keep the program.
Describe the pipeline you run every week and 5AM writes you a Python program that drives the 5am CLI — the same CLI you would type by hand. You read a structured plan before any code exists, edit the code once it does, and then either download the file and run it wherever you like, or hand it to us to run in the cloud with live logs. The output is a program, not a saved chain of blocks: a file you can read, diff, and keep.
5 GB free storage · Feature access and AI usage depend on your plan.
Managed Workflows in 2 minutes
Automate your media pipeline without writing code. Create a portrait proof sheet from a plain-language request, review and run the program, then see how album uploads can start the next run. Cloud runs and upload automation require Ultra and workflow compute credits.
Play with sound. Use the player controls for captions and full screen.
Open video in a new tab ↗Read video transcript
Automate your media pipeline with 5AM Workflows. Describe what you need in plain language. You don't need to write code. For this shoot, let's turn six portraits into a client proof sheet and a CSV index, while keeping the originals unchanged. Tell 5AM which photos to use, how to arrange them, and what files to create. Make the album an input to reuse this job. First, review the plan: the inputs, steps, outputs, and access the workflow needs. 5AM writes the Python program for you. Ask for changes in plain language, or edit and download the code. Run the built-in AI review and read its findings before starting. Here, the review flags warnings for us to check. Choose your album and run in the cloud. Reuse the same program with a different album for your next shoot. Follow the live logs to see what happened at each step. This run processed all six portraits successfully. Preview the finished proof sheet and download the files. Keep the JPEG, the CSV index, and your Python program. To automate the next batch, open Automate and choose the album to watch. New uploads can start the workflow for you. Set a quiet window so a burst of uploads becomes one run. Here, it waits five minutes after the last upload. Save the run inputs and turn it on when you're ready. Cloud runs and upload automation require Ultra and use workflow compute credits. Start with one media task you repeat. Describe the job, keep the program, and reuse it.
Why Managed Workflows?
A plan first, with its assumptions and open questions written out — thirty seconds to review, versus reading two hundred lines of generated Python
The program is yours: download workflow.py with its requirements and run it in cron, in CI, anywhere — downloading needs no subscription at all
Two gates, asking different questions — a static check that every 5am command and flag is real, and a second model reading adversarially for what will break or cost you
Runaway jobs cost exactly the ceiling you were shown: credits are reserved before a run and settled to its real duration
Each cloud run gets its own credential, scoped to what the plan declared and revoked when the run ends — never your personal token, never admin
Key Features
Ready to get started?
Start free with 5 GB — Managed Workflows is part of one hub for storing, editing, generating, and selling your media.