diff --git a/.gitignore b/.gitignore
index a7064ad40d..08f9d8ee6b 100755
--- a/.gitignore
+++ b/.gitignore
@@ -20,6 +20,7 @@ unsloth_compiled_cache/
 # ML artifacts (large files)
 outputs/
 exports/
+/datasets/
 unsloth_training_checkpoints/
 *.gguf
 *.safetensors
diff --git a/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/builder_config.json b/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/builder_config.json
deleted file mode 100644
index c204eab0fa..0000000000
--- a/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/builder_config.json
+++ /dev/null
@@ -1,72 +0,0 @@
-{
-  "data_designer": {
-    "columns": [
-      {
-        "name": "generated_instruction",
-        "drop": false,
-        "allow_resize": false,
-        "column_type": "llm-text",
-        "prompt": "Based on this target answer:\n{{ output }}\n\nWrite one high-quality plain text short and brief user instruction that this answer would satisfy.\nReturn only the instruction.",
-        "model_alias": "ministral",
-        "system_prompt": null,
-        "multi_modal_context": null,
-        "tool_alias": null,
-        "with_trace": "none",
-        "extract_reasoning_content": false
-      },
-      {
-        "name": "instruction",
-        "drop": false,
-        "allow_resize": false,
-        "column_type": "seed-dataset"
-      },
-      {
-        "name": "input",
-        "drop": false,
-        "allow_resize": false,
-        "column_type": "seed-dataset"
-      },
-      {
-        "name": "output",
-        "drop": false,
-        "allow_resize": false,
-        "column_type": "seed-dataset"
-      }
-    ],
-    "model_configs": [
-      {
-        "alias": "ministral",
-        "model": "mistralai/ministral-8b-2512",
-        "inference_parameters": {
-          "generation_type": "chat-completion",
-          "max_parallel_requests": 4,
-          "timeout": null,
-          "extra_body": null,
-          "temperature": 0.7,
-          "top_p": null,
-          "max_tokens": 1024
-        },
-        "provider": "openai_provider",
-        "skip_health_check": false
-      }
-    ],
-    "tool_configs": [],
-    "seed_config": {
-      "source": {
-        "seed_type": "hf",
-        "path": "datasets/unsloth/alpaca-cleaned/**/*.json",
-        "token": null,
-        "endpoint": "https://huggingface.co"
-      },
-      "sampling_strategy": "ordered",
-      "selection_strategy": {
-        "start": 1,
-        "end": 100
-      }
-    },
-    "constraints": null,
-    "profilers": null,
-    "processors": null
-  },
-  "library_version": "0.5.1"
-}
\ No newline at end of file
diff --git a/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/metadata.json b/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/metadata.json
deleted file mode 100644
index 4b84f8d0f4..0000000000
--- a/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/metadata.json
+++ /dev/null
@@ -1,36 +0,0 @@
-{
-  "actual_num_records": 50,
-  "buffer_size": 50,
-  "column_statistics": [
-    {
-      "column_name": "generated_instruction",
-      "column_type": "llm-text",
-      "input_tokens_mean": 170.52,
-      "input_tokens_median": 132.5,
-      "input_tokens_stddev": 119.91,
-      "num_null": 0,
-      "num_records": 50,
-      "num_unique": 50,
-      "output_tokens_mean": 39.74,
-      "output_tokens_median": 31.0,
-      "output_tokens_stddev": 39.58,
-      "pyarrow_dtype": "string",
-      "simple_dtype": "string"
-    }
-  ],
-  "dataset_name": "recipe_a28d33f9eab14660bba2437a6dfda290",
-  "file_paths": {
-    "parquet-files": [
-      "parquet-files/batch_00000.parquet"
-    ]
-  },
-  "num_completed_batches": 1,
-  "schema": {
-    "generated_instruction": "string",
-    "input": "string",
-    "instruction": "string",
-    "output": "string"
-  },
-  "target_num_records": 50,
-  "total_num_batches": 1
-}
\ No newline at end of file
diff --git a/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/parquet-files/batch_00000.parquet b/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/parquet-files/batch_00000.parquet
deleted file mode 100644
index 06eebafa26..0000000000
Binary files a/datasets/recipes/recipe_a28d33f9eab14660bba2437a6dfda290/parquet-files/batch_00000.parquet and /dev/null differ
diff --git a/studio/backend/core/data_recipe/jobs/worker.py b/studio/backend/core/data_recipe/jobs/worker.py
index 2d33a1c632..ac27cd0d0b 100644
--- a/studio/backend/core/data_recipe/jobs/worker.py
+++ b/studio/backend/core/data_recipe/jobs/worker.py
@@ -10,7 +10,7 @@ from typing import Any
 from ..service import build_config_builder, create_data_designer
 
 _PROJECT_ROOT = Path(__file__).resolve().parents[5]
-_ARTIFACT_ROOT = _PROJECT_ROOT / "datasets" / "recipes"
+_ARTIFACT_ROOT = _PROJECT_ROOT / "studio" / "backend" / "assets" / "datasets"
 
 
 class _QueueLogHandler(logging.Handler):