diff --git a/.github/reusable-build/action.yml b/.github/reusable-build/action.yml
new file mode 100644
index 00000000000..3c8b4d2839a
--- /dev/null
+++ b/.github/reusable-build/action.yml
@@ -0,0 +1,64 @@
+name: Reusable action which builds serving
+
+inputs:
+  python-version:
+    description: 'Python version'
+    required: true
+env:
+  USE_BAZEL_VERSION: '6.5.0'
+runs:
+  using: 'composite'
+  steps:
+    - name: Build the package for Python ${{ inputs.python-version }}
+      shell: bash
+      run: |
+        docker build -f tensorflow_serving/tools/docker/Dockerfile.devel
+
+    # - name: Install runner dependencies
+    #   shell: bash
+    #   run: |
+    #     sudo apt update
+    #     sudo apt install -y automake build-essential ca-certificates curl git gcc-10 g++-10 libcurl3-dev libfreetype6-dev libpng-dev libtool libzmq3-dev mlocate openjdk-8-jdk openjdk-8-jre-headless pkg-config software-properties-common swig unzip wget zip zlib1g-dev
+    #
+    # - name: Install python dependencies
+    #   shell: bash
+    #   run: |
+    #     # This uses the pre-PEP517 setuptools approach, which means
+    #     # wheels aren't built in isolated environments. We therefore need to install
+    #     # a bunch of build-time stuff here.
+    #     pip install future>=0.17.1 grpcio h5py keras_applications>=1.0.8 keras_preprocessing>=1.1.0 mock numpy portpicker requests
+    #
+    # - uses: bazel-contrib/setup-bazel@0.8.5
+    #   name: Set up Bazel
+    #   with:
+    #     # Avoid downloading Bazel every time.
+    #     bazelisk-cache: true
+    #     # Store build cache per workflow.
+    #     disk-cache: ${{ github.workflow }}-${{ hashFiles('.github/workflows/wheels.yml') }}
+    #     # Share repository cache between workflows.
+    #     repository-cache: true
+    #
+    # - name: Verify bazel installation
+    #   shell: bash
+    #   run: |
+    #     which bazel
+    #     bazel info
+    #     bazel version
+    #
+    # - name: Build the package with bazel
+    #   shell: bash
+    #   run: |
+    #     bazel build --output_filter=DONT_MATCH_ANYTHING --config=kokoro tensorflow_serving/model_servers:tensorflow_model_server
+    #     # cp bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server /usr/local/bin
+    #
+    #     # Build and install TensorFlow Serving API
+    #     bazel build --output_filter=DONT_MATCH_ANYTHING --config=kokoro tensorflow_serving/tools/pip_package:build_pip_package
+    #     bazel-bin/tensorflow_serving/tools/pip_package/build_pip_package /tmp/pip
+    #     cp /tmp/pip/tensorflow_serving_api-*.whl ./dist/
+    #     # pip --no-cache-dir install --upgrade /tmp/pip/tensorflow_serving_api-*.whl
+
+    - name: Upload wheel artifact for Python ${{ inputs.python-version }}
+      uses: actions/upload-artifact@v4
+      with:
+        name: ml-metadata-wheel-py${{ inputs.python-version }}
+        path: dist/tensorflow_serving_api-*.whl
diff --git a/.github/workflows/wheels.yml b/.github/workflows/wheels.yml
new file mode 100644
index 00000000000..6d946a2cf83
--- /dev/null
+++ b/.github/workflows/wheels.yml
@@ -0,0 +1,65 @@
+name: Build serving wheels
+
+on:
+  pull_request:
+  workflow_dispatch:
+  release:
+    types: [published]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.9", "3.10", "3.11"]
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Build package
+        id: build-package
+        uses: ./.github/reusable-build
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: List and check wheels
+        shell: bash
+        run: |
+          # Pin transitive dependency on pkginfo until https://github.com/pypa/twine/issues/1070 is
+          # fixed
+          pip install twine pkginfo>=1.10.0
+          ${{ matrix.ls || 'ls -lh' }} wheels/
+          twine check wheels/*
+
+
+  upload_to_pypi:
+    name: Upload to PyPI
+    runs-on: ubuntu-latest
+    if: (github.event_name == 'release' && startsWith(github.ref, 'refs/tags')) || (github.event_name == 'workflow_dispatch')
+    needs: [build]
+    environment:
+      name: pypi
+      url: https://pypi.org/p/tensorflow_serving/
+    permissions:
+      id-token: write
+    steps:
+      - name: Retrieve wheels
+        uses: actions/download-artifact@v4
+        with:
+          merge-multiple: true
+          path: wheels
+
+      - name: List the build artifacts
+        run: |
+          ls -lAs wheels/
+
+      - name: Upload to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          packages_dir: wheels/
diff --git a/.gitignore b/.gitignore
index fd22d60859d..82f927558a3 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,14 +1,162 @@
-node_modules
-/bazel-bin
-/bazel-ci_build-cache
-/bazel-genfiles
-/bazel-out
-/bazel-serving
-/bazel-tensorflow
-/bazel-tensorflow_serving
-/bazel-testlogs
-/bazel-tf
-/bazel-workspace
-/third_party/py/numpy/numpy_include
-/util/python/python_include
-/util/python/python_lib
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/latest/usage/project/#working-with-version-control
+.pdm.toml
+.pdm-python
+.pdm-build/
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
diff --git a/docker-compose.yml b/docker-compose.yml
new file mode 100644
index 00000000000..68b4ff0f54c
--- /dev/null
+++ b/docker-compose.yml
@@ -0,0 +1,10 @@
+services:
+  build-wheels:
+    build:
+      context: .
+      dockerfile: tensorflow_serving/tools/docker/Dockerfile.devel
+      args:
+        - PYTHON_VERSION=${PYTHON_VERSION}
+        - USE_BAZEL_VERSION=${USE_BAZEL_VERSION}
+    volumes:
+      - .:/build:delegated
diff --git a/tensorflow_serving/tools/docker/Dockerfile.devel b/tensorflow_serving/tools/docker/Dockerfile.devel
index fe28614b718..a9a9b1edc03 100644
--- a/tensorflow_serving/tools/docker/Dockerfile.devel
+++ b/tensorflow_serving/tools/docker/Dockerfile.devel
@@ -11,17 +11,16 @@
 # 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.
-FROM ubuntu:20.04 as base_build
+# FROM ubuntu:20.04 as base_build
+FROM quay.io/condaforge/miniforge3
+LABEL maintainer=gvasudevan@google.com
 
-ARG TF_SERVING_VERSION_GIT_BRANCH=master
-ARG TF_SERVING_VERSION_GIT_COMMIT=HEAD
-ARG DEBIAN_FRONTEND=noninteractive
+ENV DEBIAN_FRONTEND=noninteractive
 
-LABEL maintainer=gvasudevan@google.com
-LABEL tensorflow_serving_github_branchtag=${TF_SERVING_VERSION_GIT_BRANCH}
-LABEL tensorflow_serving_github_commit=${TF_SERVING_VERSION_GIT_COMMIT}
+ARG PYTHON_VERSION
+ARG USE_BAZEL_VERSION
 
-RUN apt-get update && apt-get install -y --no-install-recommends \
+RUN apt update -y && apt install -y --no-install-recommends \
         automake \
         build-essential \
         ca-certificates \
@@ -38,96 +37,46 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         openjdk-8-jdk\
         openjdk-8-jre-headless \
         pkg-config \
-        python-dev \
         software-properties-common \
         swig \
         unzip \
         wget \
         zip \
-        zlib1g-dev \
-        python3-distutils \
-        && \
-    apt-get clean && \
-    rm -rf /var/lib/apt/lists/*
-
-# Install python 3.9.
-RUN add-apt-repository ppa:deadsnakes/ppa && \
-    apt-get update && apt-get install -y \
-    python3.9 python3.9-dev python3-pip python3.9-venv && \
-    rm -rf /var/lib/apt/lists/* && \
-    python3.9 -m pip install pip --upgrade && \
-    update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.9 0
-
-# Make python3.9 the default python version
-# If updating default python version, also update version in the WORKSPACE file.
-RUN update-alternatives --install /usr/bin/python python /usr/bin/python3.9 0
+        zlib1g-dev
 
-RUN curl -fSsL -O https://bootstrap.pypa.io/get-pip.py && \
-    python3 get-pip.py && \
-    rm get-pip.py
+# Install python
+RUN conda create -n build -y python==${PYTHON_VERSION}
+RUN conda run -n build pip install future>=0.17.1 grpcio h5py keras_applications>=1.0.8 keras_preprocessing>=1.1.0 mock numpy portpicker requests
 
-RUN pip3 --no-cache-dir install \
-    future>=0.17.1 \
-    grpcio \
-    h5py \
-    keras_applications>=1.0.8 \
-    keras_preprocessing>=1.1.0 \
-    mock \
-    numpy \
-    portpicker \
-    requests \
-    --ignore-installed setuptools \
-    --ignore-installed six>=1.12.0
+# Set up bazelisk v1.22.0
+RUN wget https://github.com/bazelbuild/bazelisk/releases/download/v1.22.0/bazelisk-linux-amd64 -O /usr/local/bin/bazel && chmod +x /usr/local/bin/bazel
 
-# Set up Bazel
-ENV BAZEL_VERSION 6.5.0
-WORKDIR /
-RUN mkdir /bazel && \
-    cd /bazel && \
-    curl -H "User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/57.0.2987.133 Safari/537.36" -fSsL -O https://github.com/bazelbuild/bazel/releases/download/$BAZEL_VERSION/bazel-$BAZEL_VERSION-installer-linux-x86_64.sh && \
-    curl -H "User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/57.0.2987.133 Safari/537.36" -fSsL -o /bazel/LICENSE.txt https://raw.githubusercontent.com/bazelbuild/bazel/master/LICENSE && \
-    chmod +x bazel-*.sh && \
-    ./bazel-$BAZEL_VERSION-installer-linux-x86_64.sh && \
-    cd / && \
-    rm -f /bazel/bazel-$BAZEL_VERSION-installer-linux-x86_64.sh
+# Copy the repo into the container
+COPY . /build
+WORKDIR /build
 
-# Download TF Serving sources (optionally at specific commit).
-WORKDIR /tensorflow-serving
-RUN curl -sSL --retry 5 https://github.com/tensorflow/serving/tarball/${TF_SERVING_VERSION_GIT_COMMIT} | tar --strip-components=1 -xzf -
-
-FROM base_build as binary_build
+# FROM base_build as binary_build
 # Build, and install TensorFlow Serving
 ARG TF_SERVING_BUILD_OPTIONS="--config=release"
 RUN echo "Building with build options: ${TF_SERVING_BUILD_OPTIONS}"
 ARG TF_SERVING_BAZEL_OPTIONS=""
 RUN echo "Building with Bazel options: ${TF_SERVING_BAZEL_OPTIONS}"
 
-RUN bazel build --color=yes --curses=yes \
+RUN mkdir dist/
+
+RUN conda run -n build bazel build \
     ${TF_SERVING_BAZEL_OPTIONS} \
-    --verbose_failures \
     --output_filter=DONT_MATCH_ANYTHING \
     ${TF_SERVING_BUILD_OPTIONS} \
     --config=kokoro \
-    tensorflow_serving/model_servers:tensorflow_model_server && \
-    cp bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server \
-    /usr/local/bin/
+    tensorflow_serving/model_servers:tensorflow_model_server
 
 # Build and install TensorFlow Serving API
-RUN bazel build --color=yes --curses=yes \
+RUN conda run -n build bazel build \
     ${TF_SERVING_BAZEL_OPTIONS} \
-    --verbose_failures \
     --output_filter=DONT_MATCH_ANYTHING \
     ${TF_SERVING_BUILD_OPTIONS} \
     --config=kokoro \
     tensorflow_serving/tools/pip_package:build_pip_package && \
     bazel-bin/tensorflow_serving/tools/pip_package/build_pip_package \
-    /tmp/pip && \
-    pip --no-cache-dir install --upgrade \
-    /tmp/pip/tensorflow_serving_api-*.whl && \
-    rm -rf /tmp/pip
-
-FROM binary_build as clean_build
-# Clean up Bazel cache when done.
-RUN bazel clean --expunge --color=yes && \
-    rm -rf /root/.cache
-CMD ["/bin/bash"]
+    ./dist