diff --git a/.jules/bolt.md b/.jules/bolt.md
index 8780446..92dd6d8 100644
--- a/.jules/bolt.md
+++ b/.jules/bolt.md
@@ -1,3 +1,7 @@
 ## 2024-05-23 - [Regex Pre-compilation in Loops]
 **Learning:** Pre-compiling regular expressions (`re.compile`) at the module level provides a significant performance boost (measured ~1.8x speedup) when the regex is used inside a tight loop or a pandas `apply` function, compared to compiling it repeatedly or implicitly inside the loop. Vectorized string operations in Pandas are usually faster, but in complex logic cases (multiple prioritized regex groups + fallback logic), a simple pre-compiled regex with `apply` can sometimes be cleaner and sufficiently fast, or even faster if the vectorized approach requires multiple passes or expensive intermediate structures.
 **Action:** Always check for regex usage in loops or `apply` calls. If found, refactor to use module-level pre-compiled patterns. When considering vectorization, benchmark against the optimized loop version, as the overhead of complex vectorization might outweigh the benefits for moderate dataset sizes.
+
+## 2024-10-30 - [Streamlit File Upload Memory Optimization]
+**Learning:** Loading large text files (like MGF/mzTab) in Streamlit using `file.read().decode('utf-8')` creates massive memory spikes because it loads the entire raw bytes AND the decoded string into RAM before parsing. Using `io.TextIOWrapper(file, encoding='utf-8')` wraps the underlying binary stream and allows libraries like `pyteomics` to stream-read the data, reducing memory usage from O(N) to O(buffer_size).
+**Action:** Always wrap `streamlit.UploadedFile` (or any `BytesIO` source) with `io.TextIOWrapper` when passing to parsers that accept file-like objects, instead of reading and decoding the whole file first.
diff --git a/app.py b/app.py
index aa7990e..21f47b0 100644
--- a/app.py
+++ b/app.py
@@ -32,9 +32,9 @@ def run_streamlit_app():
     # Process files only when both are uploaded
     if mgf_file and mztab_file:
         # Decode uploaded file contents (Streamlit files are bytes by default)
-        # Use StringIO to create file-like objects for pyteomics parsers
-        spectra = load_mgf(io.StringIO(mgf_file.read().decode('utf-8')))
-        psm_df = load_mztab(io.StringIO(mztab_file.read().decode('utf-8')))
+        # ⚡ OPTIMIZATION: Use TextIOWrapper to stream file content instead of loading fully into memory
+        spectra = load_mgf(io.TextIOWrapper(mgf_file, encoding='utf-8'))
+        psm_df = load_mztab(io.TextIOWrapper(mztab_file, encoding='utf-8'))
 
         # Create mappings between PSMs and spectra
         mapped = map_psms_to_spectra(spectra, psm_df)
diff --git a/tests/test_streaming_io.py b/tests/test_streaming_io.py
new file mode 100644
index 0000000..307d0c4
--- /dev/null
+++ b/tests/test_streaming_io.py
@@ -0,0 +1,50 @@
+
+import io
+import unittest
+from data_loading import load_mgf, load_mztab
+
+class TestStreamingIO(unittest.TestCase):
+    """
+    Test that data loading functions support streaming input (TextIOWrapper).
+    This ensures that the memory optimization in app.py remains valid.
+    """
+    def test_load_mgf_with_textiowrapper(self):
+        mgf_content = b"""BEGIN IONS
+TITLE=Spectrum 1
+PEPMASS=1000.0
+100.0 10.0
+200.0 20.0
+END IONS
+"""
+        # Simulate Streamlit UploadedFile (BytesIO)
+        mgf_file = io.BytesIO(mgf_content)
+
+        # Wrap with TextIOWrapper
+        text_wrapper = io.TextIOWrapper(mgf_file, encoding='utf-8')
+
+        # Test load_mgf
+        spectra = load_mgf(text_wrapper)
+        self.assertEqual(len(spectra), 1)
+        self.assertEqual(spectra[0]['title'], 'Spectrum 1')
+
+    def test_load_mztab_with_textiowrapper(self):
+        mztab_content = b"""MTD\tmzTab-version\t1.0.0
+MTD\tmzTab-mode\tSummary
+MTD\tmzTab-type\tQuantification
+PSH\tsequence\tPSM_ID\taccession\tunique\tdatabase\tdatabase_version\tsearch_engine\tsearch_engine_score[1]\tmodifications\tretention_time\tcharge\texp_mass_to_charge\tcalc_mass_to_charge\tspectra_ref\tpre\tpost\tstart\tend\topt_global_cv_MS:1002217_decoy_peptide
+PSM\tPEPTIDE\tpsm1\tP12345\t0\tDB\t1.0\tMS:1001207\t100.0\t\t100.0\t2\t1000.0\t1000.0\tindex=1\t-\t-\t-\t-\t0
+"""
+        # Simulate Streamlit UploadedFile (BytesIO)
+        mztab_file = io.BytesIO(mztab_content)
+
+        # Wrap with TextIOWrapper
+        text_wrapper = io.TextIOWrapper(mztab_file, encoding='utf-8')
+
+        # Test load_mztab
+        psm_df = load_mztab(text_wrapper)
+        self.assertFalse(psm_df.empty)
+        self.assertEqual(len(psm_df), 1)
+        self.assertEqual(psm_df.iloc[0]['sequence'], 'PEPTIDE')
+
+if __name__ == '__main__':
+    unittest.main()