> ## Documentation Index
> Fetch the complete documentation index at: https://docs.btsscorp.com/llms.txt
> Use this file to discover all available pages before exploring further.

# The conversion lifecycle

> From POC to prod cutover across cycles and environments

A real conversion isn't one load — it's a series of cycles that get faster and cleaner as they move toward go-live. Each cycle is bound to one connection, and a connection is one **environment** (test → … → production), so promoting a cycle means running the next one against the next environment.

<Steps>
  <Step title="Define">
    Create the project, pick the source system, and identify the entities to convert. Plan your cycles.
  </Step>

  <Step title="POC — cycle-0">
    Run a sample conversion on a test pod: map, validate, preview, load a slice, and work the errors until it's clean.
  </Step>

  <Step title="Iterate & promote">
    Open the next cycle against the next environment's connection. **Import from cycle** carries mappings forward and learnings carry automatically. Repeat, faster each time.
  </Step>

  <Step title="Prod cutover">
    The final cycle runs against the production connection for the go-live load.
  </Step>
</Steps>

## Promoting a cycle

"Copying a cycle" means:

1. Create a new cycle bound to the **next environment's connection**.
2. Use **Import from cycle** to pull the previous cycle's mappings forward (with a diff preview).
3. The **ORCA Learnings** come with it automatically.

<Tip>
  This is why cycle-1 onward is dramatically faster than the cycle-0 POC — the mappings and every confirmed error fix are already in place.
</Tip>

<Note>
  Environments are modelled as **connections**, not a separate concept — a cycle points at whichever connection (test, UAT, prod) you're promoting into. There is no one-click "clone cycle across environments" button today; you create the next cycle and use Import from cycle.
</Note>