diff --git a/setup.py b/setup.py
new file mode 100644
index 000000000..4859ec78a
--- /dev/null
+++ b/setup.py
@@ -0,0 +1,37 @@
+from io import open
+
+from setuptools import find_packages, setup
+
+with open("requirements.txt") as f:
+    parsed_requirements = f.read().splitlines()
+# remove blank lines and comments
+parsed_requirements = [
+    x.strip()
+    for x in parsed_requirements
+    if ((x.strip()[0] != "#") and (len(x.strip()) > 3))
+]
+
+
+setup(
+    name="haystack",
+    version="0.1.0",
+    author="Malte Pietsch, Timo Moeller, Branden Chan, Tanay Soni",
+    author_email="malte.pietsch@deepset.ai",
+    description="Neural Question Answering at Scale. Use modern transformer based models like BERT to find answers in large document collections",
+    long_description=open("readme.rst", "r", encoding="utf-8").read(),
+    long_description_content_type="text/x-rst",
+    keywords="QA Question-Answering Reader Retriever BERT roberta albert squad mrc transfer-learning language-model transformer",
+    license="Apache",
+    url="https://github.com/deepset-ai/haystack",
+    download_url="https://github.com/deepset-ai/haystack/archive/0.1.0.tar.gz",
+    packages=find_packages(exclude=["*.tests", "*.tests.*", "tests.*", "tests"]),
+    install_requires=parsed_requirements,
+    python_requires=">=3.6.0",
+    tests_require=["pytest"],
+    classifiers=[
+        "Intended Audience :: Science/Research",
+        "License :: OSI Approved :: Apache Software License",
+        "Programming Language :: Python :: 3",
+        "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    ],
+)