diff --git a/RELEASE.md b/RELEASE.md
index 636ea9bdac..223da9e9ec 100644
--- a/RELEASE.md
+++ b/RELEASE.md
@@ -24,6 +24,7 @@ and this project adheres to
     sampling.
 *   `tff.program.ComputationArg`, which is helpful when creating a federated
     platform.
+*   Add `FilterToLiveChannels` to pybindings.
 
 ### Removed
 
diff --git a/tensorflow_federated/cc/core/impl/executor_stacks/executor_stack_bindings.cc b/tensorflow_federated/cc/core/impl/executor_stacks/executor_stack_bindings.cc
index cc469eb720..6f6253b4a9 100644
--- a/tensorflow_federated/cc/core/impl/executor_stacks/executor_stack_bindings.cc
+++ b/tensorflow_federated/cc/core/impl/executor_stacks/executor_stack_bindings.cc
@@ -33,6 +33,12 @@ namespace py = ::pybind11;
 namespace {
 
 PYBIND11_MODULE(executor_stack_bindings, m) {
+  m.def("filter_to_live_channels",
+        py::overload_cast<
+            const std::vector<std::shared_ptr<grpc::ChannelInterface>>&, int>(
+            &FilterToLiveChannels),
+        "Wait and filter channels that are ready or idle.");
+
   m.def("create_remote_executor_stack",
         py::overload_cast<
             const std::vector<std::shared_ptr<grpc::ChannelInterface>>&,
diff --git a/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.cc b/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.cc
index 2d6ef5a89a..430f16856e 100644
--- a/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.cc
+++ b/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.cc
@@ -44,9 +44,9 @@ namespace tensorflow_federated {
 // which are ready or idle. This function blocks up to
 // `wait_connected_duration_millis` milliseconds, awaiting all channels in
 // `channels` to be ready.
-std::vector<std::shared_ptr<grpc::ChannelInterface>> FilterToLiveChannels_(
+std::vector<std::shared_ptr<grpc::ChannelInterface>> FilterToLiveChannels(
     const std::vector<std::shared_ptr<grpc::ChannelInterface>>& channels,
-    int wait_connected_duration_millis = 1000) {
+    int wait_connected_duration_millis) {
   std::vector<std::shared_ptr<grpc::ChannelInterface>> live_channels;
   auto wait_connected =
       [&wait_connected_duration_millis](
@@ -154,7 +154,7 @@ absl::StatusOr<std::shared_ptr<Executor>> CreateRemoteExecutorStack(
   }
 
   const std::vector<std::shared_ptr<grpc::ChannelInterface>> live_channels =
-      FilterToLiveChannels_(channels);
+      FilterToLiveChannels(channels);
   if (live_channels.empty()) {
     return absl::UnavailableError(
         "No TFF workers are ready; try again to reconnect");
diff --git a/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.h b/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.h
index 2174267d07..6b5e033dba 100644
--- a/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.h
+++ b/tensorflow_federated/cc/core/impl/executor_stacks/remote_stacks.h
@@ -35,6 +35,10 @@ using ComposingChildFn = std::function<absl::StatusOr<ComposingChild>(
 using ComposingExecutorFn = std::function<std::shared_ptr<Executor>(
     std::shared_ptr<Executor>, std::vector<ComposingChild>)>;
 
+std::vector<std::shared_ptr<grpc::ChannelInterface>> FilterToLiveChannels(
+    const std::vector<std::shared_ptr<grpc::ChannelInterface>>& channels,
+    int wait_connected_duration_millis = 1000);
+
 // Creates an executor stack which proxies for a group of remote workers.
 //
 // Upon object construction, the channels which represent connections to these
diff --git a/tensorflow_federated/python/core/impl/executor_stacks/executor_stack_bindings.py b/tensorflow_federated/python/core/impl/executor_stacks/executor_stack_bindings.py
index cf145e1c31..624199e43e 100644
--- a/tensorflow_federated/python/core/impl/executor_stacks/executor_stack_bindings.py
+++ b/tensorflow_federated/python/core/impl/executor_stacks/executor_stack_bindings.py
@@ -21,6 +21,16 @@
 from tensorflow_federated.python.core.impl.types import placements
 
 
+def filter_to_live_channels(
+    channels: Sequence[executor_bindings.GRPCChannel],
+    wait_connected_duration_millis: int = 1000,
+) -> Sequence[executor_bindings.GRPCChannel]:
+  """Waits and filters channels that are ready or idle."""
+  return executor_stack_bindings.filter_to_live_channels(
+      channels, wait_connected_duration_millis
+  )
+
+
 def create_remote_executor_stack(
     channels: Sequence[executor_bindings.GRPCChannel],
     cardinalities: Mapping[placements.PlacementLiteral, int],