How to Optimize a Trading Strategy's Parameters Automatically

Every strategy has knobs — lookback lengths, thresholds, filter settings — and hand-tuning them one at a time by re-running a backtest after every guess doesn't scale past a couple of parameters.

Instead of picking one value per parameter, set a min/max/step range for each numeric parameter you want to search. The tool computes every combination across the ranges you checked — a genuine grid search, not a random sample.

Each combination in the sweep runs its own full backtest, in parallel, up to a fixed cap per sweep so a search stays a reasonable size to actually run. Nothing here is interpolated or estimated between combinations.

Results land on a leaderboard sortable by return, profit factor, win rate, trade count, or drawdown — because the single highest-return combination is often the one with the fewest trades and the least evidence behind it. Apply the row you actually trust with one click.

Once a filter or extra indicator gets added to a strategy (by hand or via chat), its new parameters show up as their own checkable group in the Sweep panel — so a second round can search the new knobs specifically, without losing the first round's result.