diff --git a/.ci/generate_test_report_lib.py b/.ci/generate_test_report_lib.py
index 25d810f1c6d17..df95db6a1d6b0 100644
--- a/.ci/generate_test_report_lib.py
+++ b/.ci/generate_test_report_lib.py
@@ -12,6 +12,65 @@
     "https://github.com/llvm/llvm-project/issues and add the "
     "`infrastructure` label."
 )
+# The maximum number of lines to pull from a ninja failure.
+NINJA_LOG_SIZE_THRESHOLD = 500
+
+
+def _parse_ninja_log(ninja_log: list[str]) -> list[tuple[str, str]]:
+    """Parses an individual ninja log."""
+    failures = []
+    index = 0
+    while index < len(ninja_log):
+        while index < len(ninja_log) and not ninja_log[index].startswith("FAILED:"):
+            index += 1
+        if index == len(ninja_log):
+            # We hit the end of the log without finding a build failure, go to
+            # the next log.
+            return failures
+        # We are trying to parse cases like the following:
+        #
+        # [4/5] test/4.stamp
+        # FAILED: touch test/4.stamp
+        # touch test/4.stamp
+        #
+        # index will point to the line that starts with Failed:. The progress
+        # indicator is the line before this ([4/5] test/4.stamp) and contains a pretty
+        # printed version of the target being built (test/4.stamp). We use this line
+        # and remove the progress information to get a succinct name for the target.
+        failing_action = ninja_log[index - 1].split("] ")[1]
+        failure_log = []
+        while (
+            index < len(ninja_log)
+            and not ninja_log[index].startswith("[")
+            and not ninja_log[index].startswith("ninja: build stopped:")
+            and len(failure_log) < NINJA_LOG_SIZE_THRESHOLD
+        ):
+            failure_log.append(ninja_log[index])
+            index += 1
+        failures.append((failing_action, "\n".join(failure_log)))
+    return failures
+
+
+def find_failure_in_ninja_logs(ninja_logs: list[list[str]]) -> list[tuple[str, str]]:
+    """Extracts failure messages from ninja output.
+
+    This function takes stdout/stderr from ninja in the form of a list of files
+    represented as a list of lines. This function then returns tuples containing
+    the name of the target and the error message.
+
+    Args:
+      ninja_logs: A list of files in the form of a list of lines representing the log
+        files captured from ninja.
+
+    Returns:
+      A list of tuples. The first string is the name of the target that failed. The
+      second string is the error message.
+    """
+    failures = []
+    for ninja_log in ninja_logs:
+        log_failures = _parse_ninja_log(ninja_log)
+        failures.extend(log_failures)
+    return failures
 
 
 # Set size_limit to limit the byte size of the report. The default is 1MB as this
diff --git a/.ci/generate_test_report_lib_test.py b/.ci/generate_test_report_lib_test.py
index eda76ead19b9d..41f3eae591e19 100644
--- a/.ci/generate_test_report_lib_test.py
+++ b/.ci/generate_test_report_lib_test.py
@@ -19,6 +19,111 @@ def junit_from_xml(xml):
 
 
 class TestReports(unittest.TestCase):
+    def test_find_failure_ninja_logs(self):
+        failures = generate_test_report_lib.find_failure_in_ninja_logs(
+            [
+                [
+                    "[1/5] test/1.stamp",
+                    "[2/5] test/2.stamp",
+                    "[3/5] test/3.stamp",
+                    "[4/5] test/4.stamp",
+                    "FAILED: touch test/4.stamp",
+                    "Wow! This system is really broken!",
+                    "[5/5] test/5.stamp",
+                ],
+            ]
+        )
+        self.assertEqual(len(failures), 1)
+        self.assertEqual(
+            failures[0],
+            (
+                "test/4.stamp",
+                dedent(
+                    """\
+                    FAILED: touch test/4.stamp
+                    Wow! This system is really broken!"""
+                ),
+            ),
+        )
+
+    def test_no_failure_ninja_log(self):
+        failures = generate_test_report_lib.find_failure_in_ninja_logs(
+            [
+                [
+                    "[1/3] test/1.stamp",
+                    "[2/3] test/2.stamp",
+                    "[3/3] test/3.stamp",
+                ]
+            ]
+        )
+        self.assertEqual(failures, [])
+
+    def test_ninja_log_end(self):
+        failures = generate_test_report_lib.find_failure_in_ninja_logs(
+            [
+                [
+                    "[1/3] test/1.stamp",
+                    "[2/3] test/2.stamp",
+                    "[3/3] test/3.stamp",
+                    "FAILED: touch test/3.stamp",
+                    "Wow! This system is really broken!",
+                    "ninja: build stopped: subcommand failed.",
+                ]
+            ]
+        )
+        self.assertEqual(len(failures), 1)
+        self.assertEqual(
+            failures[0],
+            (
+                "test/3.stamp",
+                dedent(
+                    """\
+                    FAILED: touch test/3.stamp
+                    Wow! This system is really broken!"""
+                ),
+            ),
+        )
+
+    def test_ninja_log_multiple_failures(self):
+        failures = generate_test_report_lib.find_failure_in_ninja_logs(
+            [
+                [
+                    "[1/5] test/1.stamp",
+                    "[2/5] test/2.stamp",
+                    "FAILED: touch test/2.stamp",
+                    "Wow! This system is really broken!",
+                    "[3/5] test/3.stamp",
+                    "[4/5] test/4.stamp",
+                    "FAILED: touch test/4.stamp",
+                    "Wow! This system is maybe broken!",
+                    "[5/5] test/5.stamp",
+                ]
+            ]
+        )
+        self.assertEqual(len(failures), 2)
+        self.assertEqual(
+            failures[0],
+            (
+                "test/2.stamp",
+                dedent(
+                    """\
+                    FAILED: touch test/2.stamp
+                    Wow! This system is really broken!"""
+                ),
+            ),
+        )
+        self.assertEqual(
+            failures[1],
+            (
+                "test/4.stamp",
+                dedent(
+                    """\
+                    FAILED: touch test/4.stamp
+                    Wow! This system is maybe broken!"""
+                ),
+            ),
+        )
+
     def test_title_only(self):
         self.assertEqual(
             generate_test_report_lib.generate_report("Foo", 0, []),