diff --git a/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.test.ts b/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.test.ts
index cd14102719a21..5cc0bdad3e121 100644
--- a/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.test.ts
+++ b/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.test.ts
@@ -381,12 +381,59 @@ describe('parseMlInferenceParametersFromPipeline', () => {
       model_id: 'test-model',
       pipeline_name: 'unit-test',
       source_field: 'body',
+      field_mappings: [
+        {
+          sourceField: 'body',
+          targetField: 'ml.inference.test',
+        },
+      ],
+    });
+  });
+  it('returns pipeline parameters from ingest pipeline with multiple inference processors', () => {
+    expect(
+      parseMlInferenceParametersFromPipeline('unit-test', {
+        processors: [
+          {
+            inference: {
+              field_map: {
+                body: 'text_field',
+              },
+              model_id: 'test-model',
+              target_field: 'ml.inference.body',
+            },
+          },
+          {
+            inference: {
+              field_map: {
+                title: 'text_field',
+              },
+              model_id: 'test-model',
+              target_field: 'ml.inference.title',
+            },
+          },
+        ],
+      })
+    ).toEqual({
+      destination_field: 'body',
+      model_id: 'test-model',
+      pipeline_name: 'unit-test',
+      source_field: 'body',
+      field_mappings: [
+        {
+          sourceField: 'body',
+          targetField: 'ml.inference.body',
+        },
+        {
+          sourceField: 'title',
+          targetField: 'ml.inference.title',
+        },
+      ],
     });
   });
-  it('return null if pipeline missing inference processor', () => {
+  it('return null if pipeline is missing inference processor', () => {
     expect(parseMlInferenceParametersFromPipeline('unit-test', { processors: [] })).toBeNull();
   });
-  it('return null if pipeline missing field_map', () => {
+  it('return null if pipeline is missing field_map', () => {
     expect(
       parseMlInferenceParametersFromPipeline('unit-test', {
         processors: [
diff --git a/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.ts b/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.ts
index 3744e45b0e585..7870b22977a34 100644
--- a/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.ts
+++ b/x-pack/plugins/enterprise_search/common/ml_inference_pipeline/index.ts
@@ -216,24 +216,37 @@ export const parseMlInferenceParametersFromPipeline = (
   name: string,
   pipeline: IngestPipeline
 ): CreateMlInferencePipelineParameters | null => {
-  const processor = pipeline?.processors?.find((proc) => proc.inference !== undefined);
-  if (!processor || processor?.inference === undefined) {
+  const inferenceProcessors = pipeline?.processors
+    ?.filter((p) => p.inference)
+    .map((p) => p.inference) as IngestInferenceProcessor[];
+  if (!inferenceProcessors || inferenceProcessors.length === 0) {
     return null;
   }
-  const { inference: inferenceProcessor } = processor;
-  const sourceFields = Object.keys(inferenceProcessor.field_map ?? {});
-  const sourceField = sourceFields.length === 1 ? sourceFields[0] : null;
-  if (!sourceField) {
-    return null;
-  }
-  return {
-    destination_field: inferenceProcessor.target_field
-      ? stripMlInferencePrefix(inferenceProcessor.target_field)
-      : inferenceProcessor.target_field,
-    model_id: inferenceProcessor.model_id,
-    pipeline_name: name,
-    source_field: sourceField,
-  };
+
+  // Extract source -> target field mappings from all inference processors in pipeline
+  const fieldMappings = inferenceProcessors
+    .map((p) => {
+      const sourceFields = Object.keys(p.field_map ?? {});
+      // We assume that there is only one source field per inference processor
+      const sourceField = sourceFields.length >= 1 ? sourceFields[0] : null;
+      return {
+        sourceField,
+        targetField: p.target_field, // Prefixed target field
+      };
+    })
+    .filter((f) => f.sourceField) as FieldMapping[];
+
+  return fieldMappings.length === 0
+    ? null
+    : {
+        destination_field: fieldMappings[0].targetField // Backward compatibility - TODO: remove after multi-field selector is implemented for all inference types
+          ? stripMlInferencePrefix(fieldMappings[0].targetField)
+          : '',
+        model_id: inferenceProcessors[0].model_id,
+        pipeline_name: name,
+        source_field: fieldMappings[0].sourceField, // Backward compatibility - TODO: remove after multi-field selector is implemented for all inference types
+        field_mappings: fieldMappings,
+      };
 };
 
 export const parseModelStateFromStats = (
diff --git a/x-pack/plugins/enterprise_search/common/types/pipelines.ts b/x-pack/plugins/enterprise_search/common/types/pipelines.ts
index ea86183ce9ea8..13f5ebb01b948 100644
--- a/x-pack/plugins/enterprise_search/common/types/pipelines.ts
+++ b/x-pack/plugins/enterprise_search/common/types/pipelines.ts
@@ -80,6 +80,7 @@ export interface CreateMlInferencePipelineParameters {
   model_id: string;
   pipeline_name: string;
   source_field: string;
+  field_mappings?: FieldMapping[];
 }
 
 export interface CreateMLInferencePipelineDefinition {
diff --git a/x-pack/plugins/enterprise_search/public/applications/enterprise_search_content/components/search_index/pipelines/ml_inference/ml_inference_logic.ts b/x-pack/plugins/enterprise_search/public/applications/enterprise_search_content/components/search_index/pipelines/ml_inference/ml_inference_logic.ts
index 45fc4f124cc3b..93f73f389df99 100644
--- a/x-pack/plugins/enterprise_search/public/applications/enterprise_search_content/components/search_index/pipelines/ml_inference/ml_inference_logic.ts
+++ b/x-pack/plugins/enterprise_search/public/applications/enterprise_search_content/components/search_index/pipelines/ml_inference/ml_inference_logic.ts
@@ -326,6 +326,7 @@ export const MLInferenceLogic = kea<
         modelID: params.model_id,
         pipelineName,
         sourceField: params.source_field,
+        fieldMappings: params.field_mappings,
       });
     },
     setIndexName: ({ indexName }) => {