diff --git a/src/sparkctl/compute_interface_factory.py b/src/sparkctl/compute_interface_factory.py
index 886e48a..90a2986 100644
--- a/src/sparkctl/compute_interface_factory.py
+++ b/src/sparkctl/compute_interface_factory.py
@@ -1,4 +1,5 @@
 from sparkctl.compute_interface import ComputeInterface
+from sparkctl.fake_compute import FakeCompute
 from sparkctl.models import ComputeEnvironment, SparkConfig
 from sparkctl.native_compute import NativeCompute
 from sparkctl.slurm_compute import SlurmCompute
@@ -11,6 +12,8 @@ def make_compute_interface(config: SparkConfig) -> ComputeInterface:
             return SlurmCompute(config)
         case ComputeEnvironment.NATIVE:
             return NativeCompute(config)
+        case ComputeEnvironment.FAKE:
+            return FakeCompute(config)
         case _:
             msg = f"{config.environment=} is not supported"
             raise NotImplementedError(msg)
diff --git a/src/sparkctl/fake_compute.py b/src/sparkctl/fake_compute.py
new file mode 100644
index 0000000..d9b6513
--- /dev/null
+++ b/src/sparkctl/fake_compute.py
@@ -0,0 +1,41 @@
+import tempfile
+from pathlib import Path
+from socket import gethostname
+
+from sparkctl.compute_interface import ComputeInterface
+
+
+class FakeCompute(ComputeInterface):
+    """Provides interface to Slurm."""
+
+    def get_node_memory_overhead_gb(
+        self, driver_memory_gb: int, node_memory_overhead_gb: int
+    ) -> int:
+        return driver_memory_gb + self._config.runtime.node_memory_overhead_gb
+
+    def get_num_workers(self) -> int:
+        return 1
+
+    def get_node_names(self) -> list[str]:
+        return [gethostname()]
+
+    def get_scratch_dir(self) -> Path:
+        return Path(tempfile.gettempdir())
+
+    def get_worker_node_names(self) -> list[str]:
+        return self.get_node_names()
+
+    def get_worker_memory_gb(self) -> int:
+        return 32
+
+    def get_worker_num_cpus(self) -> int:
+        return 12
+
+    def is_heterogeneous_slurm_job(self) -> bool:
+        return False
+
+    def is_worker_node(self, node_name: str) -> bool:
+        return True
+
+    def run_checks(self) -> None:
+        pass
diff --git a/src/sparkctl/models.py b/src/sparkctl/models.py
index 7e0d90a..59e21c1 100644
--- a/src/sparkctl/models.py
+++ b/src/sparkctl/models.py
@@ -197,6 +197,7 @@ class ComputeEnvironment(StrEnum):
     NATIVE = "native"
     # sparkctl detects workers through Slurm environment variables.
     SLURM = "slurm"
+    FAKE = "fake"
 
 
 class PostgresScripts(SparkctlBaseModel):
diff --git a/tests/conftest.py b/tests/conftest.py
index 162129f..fd303fa 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -139,7 +139,7 @@ def extract_tarball(src_file: Path, extract_dir: Path) -> None:
 
 @pytest.fixture
 def setup_local_env(tmp_path):
-    config = _create_default_config(SPARK_DIR, JAVA_DIR, tmp_path, ComputeEnvironment.NATIVE)
+    config = _create_default_config(SPARK_DIR, JAVA_DIR, tmp_path, ComputeEnvironment.FAKE)
     config.binaries.spark_path = SPARK_DIR
     match sys.platform:
         case "linux":