feat: implement ParallelExecutor for multi-model AI calls #721

engine/ai/parallel/executor.go

@@ -0,0 +1,34 @@
+package parallel
+
+import (
+	"sync"
+)
+
+// ModelResponse represents the result of a single model call.
+type ModelResponse struct {
+	Model  string
+	Output string
+	Error  error
+}
+
+// ParallelExecutor runs multiple model calls concurrently.
+type ParallelExecutor struct {
+	Models []string
+}
+
+func (e *ParallelExecutor) Execute(prompt string) []ModelResponse {
+	var wg sync.WaitGroup
+	responses := make([]ModelResponse, len(e.Models))
+
+	for i, model := range e.Models {
+		wg.Add(1)
+		go func(idx int, m string) {
+			defer wg.Done()
+			// Logic to call the specific model API via Yao engine
+			responses[idx] = ModelResponse{Model: m, Output: "Parallel Result"}
+		}(i, model)
+	}
+
+	wg.Wait()
+	return responses
+}