diff --git a/.github/workflows/unit-tests.yaml b/.github/workflows/unit-tests.yaml
new file mode 100644
index 0000000..dae1ac7
--- /dev/null
+++ b/.github/workflows/unit-tests.yaml
@@ -0,0 +1,25 @@
+name: Unit tests
+on:
+  push:
+  pull_request:
+
+jobs:
+  run-unit-tests:
+    name: Run unit tests
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out code
+        uses: actions/checkout@v3
+
+      - name: Set up python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.10'
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+
+      - name: Run unit tests
+        run: |
+          python -m unittest process_report/tests/unit_tests.py
diff --git a/process_report/__init__.py b/process_report/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/process_report.py b/process_report/process_report.py
similarity index 100%
rename from process_report.py
rename to process_report/process_report.py
diff --git a/process_report/tests/__init__.py b/process_report/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/process_report/tests/unit_tests.py b/process_report/tests/unit_tests.py
new file mode 100644
index 0000000..46843e3
--- /dev/null
+++ b/process_report/tests/unit_tests.py
@@ -0,0 +1,110 @@
+from unittest import TestCase
+import tempfile
+import pandas
+import os
+from textwrap import dedent
+from process_report import process_report
+
+class TestGetInvoiceDate(TestCase):
+    def test_get_invoice_date(self):
+        # The month in sample data is not the same
+        data = {'Invoice Month': ['2023-01', '2023-02', '2023-03']}
+        dataframe = pandas.DataFrame(data)
+
+        invoice_date = process_report.get_invoice_date(dataframe)
+
+        self.assertIsInstance(invoice_date, pandas.Timestamp)
+
+        # Assert that the invoice_date is the first item
+        expected_date = pandas.Timestamp('2023-01')
+        self.assertEqual(invoice_date, expected_date)
+
+
+class TestTimedProjects(TestCase):
+    def setUp(self):
+
+        # Without the dedent method, our data will have leading spaces which
+        # messes up the first key. Also the '\' is imporant to ignore the first
+        # new line we added so it's more readable in code.
+        self.csv_data = dedent("""\
+        Project,Start Date,End Date
+        ProjectA,2022-09,2023-08
+        ProjectB,2022-09,2023-09
+        ProjectC,2023-09,2024-08
+        ProjectD,2022-09,2024-08
+        """)
+        self.invoice_date = pandas.Timestamp('2023-09')
+
+        self.csv_file = tempfile.NamedTemporaryFile(delete=False, mode='w')
+        self.csv_file.write(self.csv_data)
+        self.csv_file.close()
+
+    def tearDown(self):
+        os.remove(self.csv_file.name)
+
+    def test_timed_projects(self):
+        excluded_projects = process_report.timed_projects(self.csv_file.name, self.invoice_date)
+
+        expected_projects = ['ProjectB', 'ProjectC', 'ProjectD']
+        self.assertEqual(excluded_projects, expected_projects)
+
+
+class TestRemoveNonBillables(TestCase):
+    def setUp(self):
+
+        data = {
+            'Manager (PI)': ['PI1', 'PI2', 'PI3', 'PI4'],
+            'Project - Allocation': ['ProjectA', 'ProjectB', 'ProjectC', 'ProjectD'],
+            'Untouch Data Column': ['DataA', 'DataB', 'DataC', 'DataD']
+        }
+        self.dataframe = pandas.DataFrame(data)
+
+        self.pi_to_exclude = ['PI2', 'PI3']
+        self.projects_to_exclude = ['ProjectB', 'ProjectC']
+
+        self.output_file = tempfile.NamedTemporaryFile(delete=False)
+
+    def tearDown(self):
+        os.remove(self.output_file.name)
+
+    def test_remove_non_billables(self):
+        process_report.remove_non_billables(self.dataframe, self.pi_to_exclude, self.projects_to_exclude, self.output_file.name)
+
+        result_df = pandas.read_csv(self.output_file.name)
+
+        self.assertNotIn('PI2', result_df['Manager (PI)'].tolist())
+        self.assertNotIn('PI3', result_df['Manager (PI)'].tolist())
+        self.assertNotIn('ProjectB', result_df['Project - Allocation'].tolist())
+        self.assertNotIn('ProjectC', result_df['Project - Allocation'].tolist())
+
+
+class TestMergeCSV(TestCase):
+    def setUp(self):
+        self.header = ['ID', 'Name', 'Age']
+        self.data = [
+            [1, 'Alice', 25],
+            [2, 'Bob', 30],
+            [3, 'Charlie', 28],
+        ]
+
+        self.csv_files = []
+
+        for _ in range(3):
+            csv_file = tempfile.NamedTemporaryFile(delete=False, mode='w', suffix='.csv')
+            self.csv_files.append(csv_file)
+            dataframe = pandas.DataFrame(self.data, columns=self.header)
+            dataframe.to_csv(csv_file, index=False)
+            csv_file.close()
+
+    def tearDown(self):
+        for csv_file in self.csv_files:
+            os.remove(csv_file.name)
+
+    def test_merge_csv(self):
+        merged_dataframe = process_report.merge_csv([csv_file.name for csv_file in self.csv_files])
+
+        expected_rows = len(self.data) * 3
+        self.assertEqual(len(merged_dataframe), expected_rows) # `len` for a pandas dataframe excludes the header row
+
+        # Assert that the headers in the merged DataFrame match the expected headers
+        self.assertListEqual(merged_dataframe.columns.tolist(), self.header)