Predict protocol V2 proposal (kubeflow#709)

* Predict protocol V2 proposal

* Add clarifying statement about use of platform

* Model version should be specified as a string

* Add endorsements

* Fix typo

* Update docs/predict-api/v2/required_api.md

Co-Authored-By: Animesh Singh <singhan@us.ibm.com>

* Update required_api.md

Co-authored-by: Animesh Singh <singhan@us.ibm.com>

docs/README.md

@@ -15,14 +15,18 @@ The InferenceService Data Plane architecture consists of a static graph of compo
 **Transformer**: The transformer enables users to define a pre and post processing step before the prediction and explanation workflows. Like the explainer, it is configured with relevant environment variables too. For common use cases, KFServing provides out-of-the-box transformers like Feast.
 
 # Data Plane (V1)
-KFServing has a standardized prediction workflow across all model frameworks. 
-
-Note: We are actively developing a V2 data plane protocol to improve performance (i.e. GRPC).
+KFServing has a standardized prediction workflow across all model frameworks.
 
 ## Predict
-All InferenceServices speak the Tensorflow V1 HTTP API: https://www.tensorflow.org/tfx/serving/api_rest#predict_api. 
+All InferenceServices speak the Tensorflow V1 HTTP API: https://www.tensorflow.org/tfx/serving/api_rest#predict_api.
 
 Note: Only Tensorflow models support the fields "signature_name" and "inputs".
 
 ## Explain
 All InferenceServices that are deployed with an Explainer support a standardized explanation API. This interface is identical to the Tensorflow V1 HTTP API with the addition of an ":explain" verb.
+
+# Data Plane (V2)
+The second version of the data-plane protocol addresses several issues found with the V1 data-plane protocol, including performance and generality across a large number of model frameworks and servers.
+
+## Predict
+The V2 protocol proposes both HTTP/REST and GRPC APIs. See the [complete proposal](/docs/predict-api/v2) for more information.

docs/predict-api/v2/README.md

@@ -0,0 +1,19 @@
+# Predict Protocol - Version 2
+
+The *Predict Protocol, version 2* is a set of HTTP/REST and GRPC APIs
+for inference / prediction servers. By implementing this protocol both
+inference clients and servers will increase their utility and
+portability by being able to operate seamlessly on platforms that have
+standardized around this protocol.
+
+The protocol is composed of a required set of APIs that must be
+implemented by a compliant server. This required set of APIs is
+described in [required_api.md](./required_api.md).
+
+The protocol supports an extension mechanism as a required part of the
+API, but no specific extensions are required to be implemented by a
+compliant server.
+
+The protocol is not yet finalized and so feedback is welcome. To
+provide feedback open an issue and prepend the title with "[Predict
+Protocol V2]".