diff --git a/hdf5_linkml/src/hdf5_linkml/nwb.py b/hdf5_linkml/src/hdf5_linkml/nwb.py
new file mode 100644
index 0000000..87e9866
--- /dev/null
+++ b/hdf5_linkml/src/hdf5_linkml/nwb.py
@@ -0,0 +1,68 @@
+import pdb
+from pathlib import Path
+import h5py
+import json
+from hdf5_linkml.io import H5File
+from nwb_linkml.adapters import NamespacesAdapter
+from nwb_linkml.io import load_schema_file
+from linkml_runtime.linkml_model import SchemaDefinition
+
+DATASET = '/Users/jonny/Dropbox/lab/p2p_ld/data/nwb/sub-738651046_ses-760693773.nwb'
+
+
+class NWBH5File(H5File):
+
+    def __init__(self, file: Path):
+        self.file = Path(file)
+
+    def read(self):
+        with h5py.File(self.file, 'r') as h5f:
+            embedded_schema = self.load_embedded_schema(h5f)
+            translated_schema = self.translate_schema(embedded_schema)
+            pdb.set_trace()
+
+    def load_embedded_schema(self, h5f:h5py.File) -> List[dict]:
+        """
+        Stored in ``/specfications/{namespace}/{version}/{schema_files}``
+
+        Args:
+            h5f:
+
+        Returns:
+            [
+                {
+                    'namespace': "DECODED_JSON_OBJECT",
+                    'nwb.base': "DECODED_JSON_OBJECT",
+                    ...
+                },
+                {
+                    'namespace': "DECODED_JSON_OBJECT",
+                }
+            ]
+
+        """
+
+        namespaces = []
+        for ns_name, ns in h5f['specifications'].items():
+            ns_schema = {}
+            for version in ns.values():
+                for schema_name, schema in version.items():
+                    # read the source json binary string
+                    sch_str = schema[()]
+                    sch_dict = json.loads(sch_str)
+                    ns_schema[schema_name] = sch_dict
+            namespaces.append(ns_schema)
+        return namespaces
+
+    def translate_schema(self, schema:List[dict]) -> List[SchemaDefinition]:
+        translated = []
+        for ns_schema in schema:
+            # find namespace
+            namespace = ns_schema['namespace']
+            ns_schema_adapters = [load_schema_file(sch_name, schema_dict) for sch_name, schema_dict in ns_schema.items() if sch_name != 'namespace']
+            adapter = NamespacesAdapter(
+                namespaces=namespace,
+                schemas=ns_schema_adapters
+            )
+            res = adapter.build()
+            translated.extend(res.schema)
\ No newline at end of file