What it is

The Workspace is where you build your strategy on a canvas in a way you can easily read. You place nodes (steps) on a canvas, connect them in a sequence, and then run a workspace backtest as one combined run.

As you create more reports, policies, and simulations over time, it can get confusing to remember where each part came from. Workspace solves that by keeping your work in one ordered visual trail: it shows the next step and makes it easy to retrace what you’ve already done.

In the UI, you always start from an account canvas (one account’s work at a time). When you connect another account to the same strategy, the Workspace lets you review both accounts’ work in a single ordered view-so you can decide where to continue investigating.

Workspace also supports executing accounts. You can assign account context to your strategy and saved policy versions, then review results by that account context so comparisons stay clear and traceable.

What you can do

  • Create and update a workspace with a clear node sequence you can retrace later
  • Connect nodes that originate from your report and simulator work (reports → policies → workspace nodes)
  • Assign executing accounts so you can compare how the same workspace logic plays out per account
  • Combine policy inputs into a single workspace backtest, including inputs from different reports/cohorts, candle timeframes, or policy types
  • Run the workspace backtest to compare baseline vs your connected setup

How it works

1. Start from your strategy context

Workspace work is tied to your selected strategy. You’ll typically begin with strategy-related inputs you already generated in reports and policy tools.

2. Build the sequence with nodes

Nodes represent the steps in your workspace canvas setup. As you add nodes, you control the order they apply in the workspace.

At a high level, you’re assembling:

  • A starting point (what trade cohort/context you’re working from)
  • One or more policy changes (the “edits” you want to test)
  • The resulting workspace backtest output you want to inspect

3. Set executing accounts (account context for comparisons)

Assign executing accounts so each saved policy setup is tied to a clear account context. This helps you verify whether your connected workspace setup improves outcomes for that context, and compare results across contexts without losing your work trail.

Once you’ve built an ordered workspace for your selected executing accounts, you can return to that setup as a clean “work trail” while you create new reports and policies off it.

4. Run the workspace backtest (counterfactual)

A workspace backtest is the single combined backtest run produced by your connected workspace setup. It applies the policies you connected across your selected inputs (for example: same-report policies, different cohort inputs, different candle timeframes, and different policy types) in one easy-to-follow result.

Use the output to compare what changes in your connected setup (not as a guarantee about future live execution).

Limitations

  • Workspace backtests are still counterfactual: they explain “what would have happened on the past cohort” rather than predicting future behavior.
  • Workspace does not model live strategy execution bar-by-bar or tick-by-tick order fills. Results reflect the same stored-trade / candle-based assumptions used in reports and the simulator.
  • Results depend on market data coverage for the underlying cohorts and inputs. If coverage is weak, labels can be unknown or results should be treated cautiously.
  • Mixing inputs from different timeframes and policy types can be useful, but it can also make comparisons less direct. Always label what each node’s input came from.