diff --git a/README.md b/README.md
new file mode 100644
index 0000000..bd5ea02
--- /dev/null
+++ b/README.md
@@ -0,0 +1,44 @@
+# Minerva
+Minerva is the Roman equivalent of Athena, and Athena is AWS's database that
+stores results in S3.
+
+In order to ease programmatic access to Athena and offer blocking access (so
+that your code waits for the result), I wrote `minerva` to make it seamless.
+
+The results are returned as pyarrow datasets (with parquet files as the
+underlying structure).
+
+# Basic Usage
+```
+import access as a
+
+athena = a.Athena("hay", "s3://haystac-pmo-athena/")
+query  = athena.query('select * from "trajectories"."kitware" limit 10')
+data   = query.results()
+print(data.head(10))
+```
+
+First, a connection to Athena is made. The first argument is the AWS profile in
+`~/.aws/credentials`. The second argument is the S3 location where the results
+will be stored.
+
+In the second substantive line, an SQL query is made. This is **non-blocking**.
+The query is off and running and you are free to do whatever you want now.
+
+In the third line, the results are requested. This is **blocking**, so the code
+will wait here (checking with AWS every 5 seconds) until the results are ready.
+Then, the results are downloaded to `/tmp/` and lazily interpreted as parquet
+files in the form of a `pyarrow.dataset.dataset`.
+
+# Returning Scalar Values
+In SQL, scalar values get assigned an anonymous column -- Athena doesn't like
+that. Thus, you have to assign the column a name.
+
+```
+data = athena.query('select count(*) as my_col from "trajectories"."kitware"').results()
+print(data.head(1))
+```
+
+# TODO
+* parallelize the downloading of files
+
diff --git a/access.py b/access.py
index 2befd90..d6c7fe4 100644
--- a/access.py
+++ b/access.py
@@ -66,6 +66,8 @@ class Query:
             # Because we're using `UNLOAD`, we get a manifest of the files
             # that make up our data.
             files   = self.manifest(tiedot).strip().split("\n")
+
+            # TODO parallelize this
             local   = [self.handler.download(f) for f in files]
             self.ds = pa.dataset.dataset(local)