diff --git a/README.rst b/README.rst
index b8868dd..6d00a7e 100644
--- a/README.rst
+++ b/README.rst
@@ -23,8 +23,7 @@ Getting started
 * To use this code, build the client using `Apache Maven`_::
 
     cd dataflow-java
-    mvn compile
-    mvn bundle:bundle
+    mvn package
 
 * Then, follow the Google Genomics `sign up instructions`_ to generate a valid
   ``client_secrets.json`` file.
@@ -36,7 +35,7 @@ Getting started
   Project ID and Google Cloud Storage bucket you made in the first step.
   This command runs the VariantSimilarity pipeline (which runs PCoA on a dataset)::
 
-    java -cp target/google-genomics-dataflow*.jar \
+    java -cp target/google-genomics-dataflow*-runnable.jar \
       com.google.cloud.genomics.dataflow.pipelines.VariantSimilarity \
       --project=your-project-id \
       --output=gs://your-bucket/output/localtest.txt \
@@ -48,7 +47,7 @@ Getting started
 * To deploy your pipeline (which runs on Google Compute Engine), some additional
   command line arguments are required::
 
-    java -cp target/google-genomics-dataflow*.jar \
+    java -cp target/google-genomics-dataflow*-runnable.jar \
       com.google.cloud.genomics.dataflow.pipelines.VariantSimilarity \
       --runner=BlockingDataflowPipelineRunner \
       --project=your-project-id \
@@ -115,7 +114,7 @@ utils:
 
 Maven artifact
 --------------
-This code is also deployed as a Maven artifact through Sonatype. The
+This code is also deployed as Maven artifacts through Sonatype, including both a normal jar and a runnable jar containing all dependencies (a fat jar). The
 `utils-java readme <https://github.com/googlegenomics/utils-java#releasing-new-versions>`_
 has detailed instructions on how to deploy new versions.
 
@@ -136,17 +135,6 @@ You can find the latest version in
 
 For an example pipeline that depends on this code in another GitHub repository, see https://github.com/googlegenomics/codelabs/tree/master/Java/PlatinumGenomes-variant-transformation.
 
-Updating the prebuilt jar
--------------------------
-
-The prebuilt jar is currently generated by hand (`but we plan to change this <https://github.com/googlegenomics/dataflow-java/issues/49>`_) whenever the code seems stable and has enough new features to warrant the effort.
-
-To update the jar, bundle and copy to a new location::
-
-  cd dataflow-java
-  mvn clean compile bundle:bundle
-  cp target/google-genomics-dataflow*.jar google-genomics-dataflow.jar
-
 Project status
 --------------
 
diff --git a/google-genomics-dataflow.jar b/google-genomics-dataflow.jar
deleted file mode 100644
index 3a302ab..0000000
Binary files a/google-genomics-dataflow.jar and /dev/null differ
diff --git a/pom.xml b/pom.xml
index fa2d101..f193210 100644
--- a/pom.xml
+++ b/pom.xml
@@ -255,17 +255,21 @@
       </plugin>
 
       <plugin>
-        <groupId>org.apache.felix</groupId>
-        <artifactId>maven-bundle-plugin</artifactId>
-        <version>2.4.0</version>
-        <extensions>true</extensions>
-        <configuration>
-          <instructions>
-            <!-- Embed all dependencies -->
-            <Embed-Transitive>true</Embed-Transitive>
-            <Embed-Dependency>*;scope=compile|runtime;inline=true</Embed-Dependency>
-          </instructions>
-        </configuration>
+        <groupId>org.apache.maven.plugins</groupId>
+        <artifactId>maven-shade-plugin</artifactId>
+        <version>2.3</version>
+        <executions>
+          <execution>
+            <phase>package</phase>
+            <goals>
+              <goal>shade</goal>
+            </goals>
+            <configuration>
+              <shadedArtifactAttached>true</shadedArtifactAttached>
+              <shadedClassifierName>runnable</shadedClassifierName>
+            </configuration>
+          </execution>
+        </executions>
       </plugin>
 
       <plugin>