diff --git a/Dockerfile b/Dockerfile
index 29eb36ac..ff5f30c9 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,3 +1,17 @@
+#    Modifications Copyright 2019 GridCell Ltd
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+
 # This Dockerfile is used to serve the AllenNLP demo.
 
 FROM allennlp/commit:31af01e0db7ac401b6c4923d5badd7de2691d6a2
@@ -5,10 +19,13 @@ LABEL maintainer="allennlp-contact@allenai.org"
 
 WORKDIR /stage/allennlp
 
+# Use cache busting to install and upgrade https://docs.docker.com/develop/develop-images/dockerfile_best-practices/
+RUN apt-get -y update && apt-get -y upgrade
+
 # Install Java.
 RUN echo "deb http://http.debian.net/debian jessie-backports main" >>/etc/apt/sources.list
-RUN apt-get update
-RUN apt-get install -y -t jessie-backports openjdk-8-jdk
+RUN apt-get update && apt-get install -y -t jessie-backports openjdk-8-jdk
+
 
 # Install npm early so layer is cached when mucking with the demo
 RUN curl -sL https://deb.nodesource.com/setup_8.x | bash - && apt-get install -y nodejs
diff --git a/NOTICE b/NOTICE
new file mode 100644
index 00000000..e3d41c5e
--- /dev/null
+++ b/NOTICE
@@ -0,0 +1,5 @@
+Modified files Copyright GridCell Ltd 2019
+
+
+This product is derived from software developed at
+AllenAI (https://allenai.org/) form repository (https://github.com/allenai/allennlp-demo).
\ No newline at end of file
diff --git a/README.md b/README.md
index 1834d27a..c6e2cc47 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,18 @@
+  <!--- 
+  Modified Copyright 2019 GridCell Ltd
+
+    Licensed under the Apache License, Version 2.0 (the "License");
+    you may not use this file except in compliance with the License.
+    You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+    Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.-->
+
 # AllenNLP Demo
 
 This repository contains the AllenNLP demo.
@@ -54,4 +69,237 @@ To run the demo for development, you will need to:
     ```
 
     Normally, the backend server would manage the frontend assets as well - the JS has a special hack for if it is running on port 3000 (which it does by default if you are running the unoptimized JS using `npm run start`), it will look for the backend service at port 8000. Otherwise, it serves the backend and the frontend from the same port.
+ 
+ # Section added by by [gridcell](https://twitter.com/gridcell_io) 
+ 
+ # Deploying on Kubernetes (Google Kubernetes Engine)
+ 
+ ## Preparation
+ 
+ 1. Docker builds
+ 
+ The script `build-allennlp-image.sh` builds the local `Dockerfile`, tags the image and uploads it to the 
+ [Google Cloud Container Registry](https://cloud.google.com/container-registry/docs/).
+ 
+ 
+ Set the following environment variables first 
+ 
+ ```
+ export PROJECT=some_project # Google Cloud project you are working under 
+ export VERSION=0.8.2        # set the VERSION to whatever you like e.g latest or 0.8.2
+ ```
+ Now set `CONTAINER_REGION`, the location of the container [registry](https://cloud.google.com/container-registry/docs/pushing-and-pulling)
+ 
+ The four options are:
+ 
+ * `gcr.io` hosts the images in the United States, but the location may change in the future
+ * `us.gcr.io` hosts the image in the United States, in a separate storage bucket from images hosted by gcr.io
+ * `eu.gcr.io` hosts the images in the European Union
+ * `asia.gcr.io` hosts the images in Asia
+  
+Example
+ 
+```bash
+export CONTAINER_REGION=us.gcr.io
+```
+
+then run it
+
+```bash
+ cd kubernetes
+./build-allennlp-image.sh
+```
+
+look out for something like 
+```bash
+The push refers to repository [us.gcr.io/PROJECT/allennlp]
+...
+DIGEST        TAGS          TIMESTAMP
+XXXXXXX  0.8.2,latest  2019-03-19T15:07:01
+XXXXXXX                2019-03-12T20:25:41
+```
+
+
+2. Kuberetes preparation
+ 
+ The `kubernetes` folder has all the kubernetes definition files to launch allennlp on a cluster
+ 
+ * `allennlp.yml` contains a [StatefulSet](https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/) that uses persistent volumes to store downloaded models on a disk attached to the node.
+  This deployment is configured with 3 node replica. 
+  
+  Be sure to change the file in the containers section and replace `CHANGE_ME` with the tag of your Docker image from 1. above:
+   i.e. `CONTAINER_REGION/PROJECT/allennlp:VERSION` e.g. `eu.gcr.io/some_project/allennlp:0.8.2`
+   **Be sure to change this to the EXACT tag you got from step 1 or else you will get errors when deploying to Kubernetes** 
+  
+   ```yml
+   containers:
+        - name: allennlp
+          image:  CHANGE_ME #enter path to docker registry e.g on google cloud container registry eu.gcr.io/PROJECT_ID/allennlp:0.8.2
+   ```
+          
+ * `al-lb-svc.yml` cretes a load balancer with a public IP address (see below)on http port 80
+ * `al-ia-svc.yml` cretes a internal (to the cluster) load balancer. For example if you have a client nodejs , python etc running on your kubernetes cluster you can refer to it by `http://al-lb-svc:8000/predict/machine-comprehension`
+ * `gce-standard-sc.yml` the storage class of the disks (you can change this to ssd if you like `type: pd-ssd`)    
+   
+  
+ ## Google Kubernetes Engine GKE
+ 
+ The AllenNLP UI and backend can be deployed to a kubernetes cluster on a given cloud platform that supports Kubernetes. Requires
+ * [Google Gloud SDK](https://cloud.google.com/sdk/install)
+ * [Kubectl](https://kubernetes.io/docs/tasks/tools/install-kubectl/) 
+ * [Minikube](https://kubernetes.io/docs/setup/minikube/) - optional if you want to test locally first
+ 
+ 1. Follow the steps [here](https://cloud.google.com/kubernetes-engine/docs/quickstart) to learn more about GKE
+ 
+ 2. Create a GKE cluster, 50GB disk, `highmem` instances are recommended for loading and serving models. Other [instances](https://cloud.google.com/compute/docs/machine-types) are available too.
+ You can also add `--preemptible` to save running costs see [here](https://cloud.google.com/kubernetes-engine/docs/how-to/preemptible-vms)  
+ 
+ ```
+ gcloud container clusters create example-cluster \
+       --node-locations us-central1-a \
+       --additional-zones us-central1-b,us-central1-c \
+       --machine-type=n1-highmem-4 \
+       --disk-size=50
+       --region us-central1
+   ```
+ 3. Get access to the cluster
+ 
+```bash
+gcloud beta container clusters get-credentials example-cluster --region us-central1 --project PROJECT
+```
+4. Apply the changes
+
+```bash
+kubectl apply -f kubernetes/allennlp.yml 
+kubectl apply -f kubernetes/es-ia-svc.yml 
+kubectl apply -f kubernetes/es-lb-svc.yml 
+kubectl apply -f kubernetes/gce-standard-sc.yml
+```
+
+5. Check deployment.
+
+Step 4 will create the stateful set with 3 persistent volumes each with its own persistent volume claim, an internal load balancer service and external load balancer
+
+```bash
+kubectl get all
+```
+
+should return something like 
+```bash
+NAME             READY     STATUS    RESTARTS   AGE
+pod/allennlp-0   1/1       Running   0          6m
+pod/allennlp-1   1/1       Running   0          4m
+pod/allennlp-2   1/1       Running   0          2m
+
+NAME                 TYPE           CLUSTER-IP     EXTERNAL-IP      PORT(S)          AGE
+service/al-ia-svc    ClusterIP      10.XX.XX.X   <none>           8000/TCP         9m
+service/al-lb-svc    LoadBalancer   10.XX.XX.X   YYY.YYY.YYY.YYY   8000:31627/TCP   9m
+service/kubernetes   ClusterIP      10.XX.XX.X    <none>           443/TCP          12m
+
+NAME                        DESIRED   CURRENT   AGE
+statefulset.apps/allennlp   3         3         6m
+
+```
+
+6. Check the API through the load balancer
+
+Using the address  `YYY.YYY.YYY.YYY` in step 5 you can query the API using Postman or Curl 
+
+`http://YYY.YYY.YYY.YYY:8000/predict/machine-comprehension`
+
+With the `POST` body
+
+```json
+{
+    "passage": "Robotics is an interdisciplinary branch of engineering and science that includes mechanical engineering,"
+    "question": "What do robots that resemble humans attempt to do?"
+}
+```
+
 
+which should return something like (shortend and trunctated):
+```json
+{
+    "best_span": [
+        147,
+        154
+    ],
+    "best_span_str": "replicate walking, lifting, speech, cognition",
+    "passage_question_attention": [
+        [
+            0.2749797999858856,
+            0.04183763265609741,
+            0.17196990549564362,
+            0.20194320380687714,
+            0.04489961266517639,
+            0.02290954254567623,
+            0.055871427059173584,
+            0.05694901570677757,
+            0.06618849188089371,
+            0.06245144084095955
+        ],
+        [
+            0.21053647994995117,
+            0.11562809348106384,
+            0.11758492887020111,
+            0.15465371310710907,
+            0.1086740493774414,
+            0.02625945955514908,
+            0.04414062201976776,
+            0.061525385826826096,
+            0.09386775642633438,
+            0.0671294555068016
+        ],
+        [
+            0.11171773821115494,
+            0.11988984048366547,
+            0.023860322311520576,
+            0.030449630692601204,
+            0.021468957886099815,
+            0.009852085262537003,
+            0.020644865930080414,
+            0.04953836649656296,
+            0.41884300112724304,
+            0.19373510777950287
+        ]
+    ],
+    "passage_tokens": [
+        "Robotics",
+        "is",
+        "an",
+        "can",
+        "do",
+        "."
+    ],
+    "question_tokens": [
+        "What",
+        "do",
+        "?"
+    ],
+    "span_end_logits": [
+        -6.517852306365967,
+        -12.051702499389648,
+        -1.8338027000427246
+    ],
+    "span_end_probs": [
+        0.0001122089524869807,
+        4.4330960236038663e-7,
+        0.009583641774952412,
+        0.07637327909469604,
+        0.048559825867414474,
+        0.01214184332638979
+    ],
+    "span_start_logits": [
+        -5.348807334899902,
+        -10.847269058227539,
+        -3.811220169067383,
+        -4.827718734741211
+    ],
+    "span_start_probs": [
+        0.000004799693670065608,
+        1.9645446158733648e-8,
+        0.000022334650566335768,
+        0.00000808201093605021
+    ]
+}
+```
\ No newline at end of file
diff --git a/kubernetes/al-ia-svc.yml b/kubernetes/al-ia-svc.yml
new file mode 100644
index 00000000..00a47d1c
--- /dev/null
+++ b/kubernetes/al-ia-svc.yml
@@ -0,0 +1,27 @@
+#  Copyright 2019 GridCell Ltd
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+
+apiVersion: v1
+kind: Service
+metadata:
+  name: al-ia-svc
+  labels:
+    app: al-ia-svc
+spec:
+  selector:
+    app: allennlp
+  ports:
+  - name: http
+    port: 8000
+    protocol: TCP
diff --git a/kubernetes/al-lb-svc.yml b/kubernetes/al-lb-svc.yml
new file mode 100644
index 00000000..dc9fff05
--- /dev/null
+++ b/kubernetes/al-lb-svc.yml
@@ -0,0 +1,28 @@
+#  Copyright 2019 GridCell Ltd
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+
+apiVersion: v1
+kind: Service
+metadata:
+  name: al-lb-svc
+  labels:
+    app: al-lb-svc
+spec:
+  type: LoadBalancer
+  selector:
+    app: allennlp
+  ports:
+  - name: http
+    port: 8000
+    protocol: TCP
\ No newline at end of file
diff --git a/kubernetes/allennlp.yml b/kubernetes/allennlp.yml
new file mode 100644
index 00000000..fb99e521
--- /dev/null
+++ b/kubernetes/allennlp.yml
@@ -0,0 +1,84 @@
+#  Copyright 2019 GridCell Ltd
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+
+apiVersion: apps/v1
+kind: StatefulSet
+metadata:
+  name: allennlp
+  labels:
+    app: allennlp
+spec:
+  selector:
+    matchLabels:
+      app: allennlp
+  # the headless-service that governs this StatefulSet
+  # responsible for the network identity of the set.
+  serviceName: allennlp-discovery-svc
+  replicas: 3
+  # Template is the object that describes the pod that will be created
+  template:
+    metadata:
+      labels:
+        app: allennlp
+    spec:
+      securityContext:
+        # allows read/write access for mounted volumes
+        # by users that belong to a group with gid: 1000
+        fsGroup: 1000
+      initContainers:
+      # init-container for setting the mmap count limit
+      - name: sysctl
+        image: busybox
+        imagePullPolicy: IfNotPresent
+        command: ["sysctl", "-w", "vm.max_map_count=262144"]
+        securityContext:
+          privileged: true
+      #optional uncomment these 2 next lines if you want allennlp to only run on specified nodes
+      #see https://kubernetes.io/docs/concepts/configuration/assign-pod-node/
+#      nodeSelector:
+#        purpose: allennlp
+      containers:
+      - name: allennlp
+        image:  CHANGE_ME #enter path to docker registry e.g on google cloud container registry eu.gcr.io/PROJECT_ID/allennlp:0.8.2
+        resources:
+            limits:
+              memory: "13Gi"
+              cpu: "2.5"
+            requests:
+              memory: "12.0Gi"
+              cpu: "2"
+        imagePullPolicy: Always
+        ports:
+        - containerPort: 8000
+          name: transport
+          protocol: TCP
+        volumeMounts:
+        - name: es-data-vc
+          mountPath: /usr/share/allennlp/data
+#  # The StatefulSet guarantees that a given [POD] network identity will
+#  # always map to the same storage identity
+  volumeClaimTemplates:
+  - metadata:
+      name: es-data-vc
+    spec:
+      accessModes: [ "ReadWriteOnce" ]
+      resources:
+        requests:
+          # elasticsearch mounted data directory size (to be adjusted based on need)
+          storage: 25Gi
+      storageClassName: gce-standard-sc
+      # no LabelSelector  defined
+      # claims can specify a label selector to further filter the set of volumes
+      # currently, a PVC with a non-empty selector can not have a PV dynamically provisioned for it
+      # no volumeName is provided
diff --git a/kubernetes/build-allennlp-image.sh b/kubernetes/build-allennlp-image.sh
new file mode 100755
index 00000000..6cbda053
--- /dev/null
+++ b/kubernetes/build-allennlp-image.sh
@@ -0,0 +1,37 @@
+#  Copyright 2019 GridCell Ltd
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+
+#!/usr/bin/env bash
+if [[ -z "${PROJECT}" ]]; then
+      echo "Please specify a Google Compute project"
+fi
+if [[ -z "${VERSION}" ]]; then
+      echo "Please specify a version"
+fi
+
+if [[ -z "${CONTAINER_REGION}" ]]; then
+      echo "Please specify a container region"
+fi
+
+
+export CWD=`pwd`
+gcloud auth configure-docker
+
+
+cd ../
+docker build . -t allennlp:$VERSION
+docker tag  allennlp:$VERSION $CONTAINER_REGION/$PROJECT/allennlp:$VERSION
+docker push $CONTAINER_REGION/$PROJECT/allennlp:$VERSION
+gcloud container images list-tags $CONTAINER_REGION/$PROJECT/allennlp
+cd $CWD
\ No newline at end of file
diff --git a/kubernetes/gce-standard-sc.yml b/kubernetes/gce-standard-sc.yml
new file mode 100644
index 00000000..86d4b8ee
--- /dev/null
+++ b/kubernetes/gce-standard-sc.yml
@@ -0,0 +1,21 @@
+#  Copyright 2019 GridCell Ltd
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+
+kind: StorageClass
+apiVersion: storage.k8s.io/v1beta1
+metadata:
+  name: gce-standard-sc
+provisioner: kubernetes.io/gce-pd
+parameters:
+  type: pd-standard