feat: Migrate dependency management from pip to uv workspaces

.python-version

@@ -0,0 +1 @@
+3.12

README.md

@@ -52,6 +52,8 @@ Kubeflow users often struggle to find relevant information across the extensive
 
 ## Prerequisites
 
+- [uv](https://docs.astral.sh/uv/) (Python package manager)
+- Python 3.9+
 - Kubernetes cluster (1.20+)
 - Helm 3.x
 - Kubeflow Pipelines
@@ -405,25 +407,47 @@ data: {"type": "done"}
 
 **Critical**: Both APIs require SSL certificates from a trusted Certificate Authority. Without proper SSL certificates, browsers will block WebSocket connections and HTTPS requests.
 
+#### Dependency Management (uv Workspaces)
+
+The API servers (`server` and `server-https`) and pipelines are managed as members of a `uv` workspace defined in the root `pyproject.toml`.
+
+- **To add a dependency to the WebSocket Server**: 
+  ```bash
+  uv add --package docs-agent-server <package-name>
+  ```
+- **To add a dependency to the HTTPS Server**: 
+  ```bash
+  uv add --package docs-agent-server-https <package-name>
+  ```
+- **Local Development**: Because they are workspace members, running `uv sync --all-packages` at the root will install dependencies for all APIs into a shared `.venv`.
+- **Building Docker Images**: The Dockerfiles use `uv` internally for blazing-fast installs. You build them from the root directory by passing the member folder as the build context:
+  ```bash
+  docker build -t docs-agent-server:latest server/
+  docker build -t docs-agent-server-https:latest server-https/
+  ```
+
 ## Usage
 
 ### Starting the Services
 
 1. **Deploy Milvus and KServe** (as described above)
 
-2. **Run the Pipeline**:
+2. **Install dependencies and activate the virtual environment**:
    ```bash
-   python pipelines/kubeflow-pipeline.py
+   uv sync --all-packages
+   source .venv/bin/activate
    ```
 
-3. **Start the API Server**:
+3. **Run the Pipeline and API Servers**:
    ```bash
-   # WebSocket API
-   python server/app.py
+   python pipelines/kubeflow-pipeline.py
 
-   # HTTPS API
-   python server-https/app.py
+   # API Servers
+   python server/app.py          # WebSocket API
+   python server-https/app.py    # HTTPS API
    ```
+
+   *(Note: You can also use `uv run python <script>` if you prefer not to activate the environment.)*
 
 ### API Usage Examples
 
@@ -681,11 +705,23 @@ We welcome contributions! Please see our [contributing guidelines](CONTRIBUTING.
 
 ### Development Setup
 
-1. Fork the repository
-2. Create a feature branch
-3. Make your changes
-4. Add tests if applicable
-5. Submit a pull request
+1. Install [uv](https://docs.astral.sh/uv/getting-started/installation/)
+2. Fork and clone the repository
+3. Install all dependencies and activate the virtual environment:
+   ```bash
+   uv sync --all-packages
+   source .venv/bin/activate
+   ```
+4. Run any script normally:
+   ```bash
+   python server/app.py
+   ```
+   *(Alternatively, use `uv run python <script>`)*
+5. To add a dependency to a specific component:
+   ```bash
+   uv add --package docs-agent-server some-package
+   ```
+6. Create a feature branch, make your changes, and submit a pull request
 
 ## License
 

pipelines/README.md

@@ -35,9 +35,12 @@ The pipelines download documentation from GitHub repositories, process the conte
 
 ### Usage
 
+First, ensure your workspace virtual environment is activated (`source .venv/bin/activate`), then run:
+
 ```bash
-python kubeflow-pipeline.py
+python pipelines/kubeflow-pipeline.py
 ```
+*(Alternatively, use `uv run python pipelines/kubeflow-pipeline.py`)*
 
 ### Parameters
 
@@ -369,6 +372,7 @@ print(f"Debug: Processing {len(file_paths_list)} files")
 
 ## Requirements
 
+- [uv](https://docs.astral.sh/uv/) for dependency management (dependencies defined in `pyproject.toml`)
 - Kubeflow Pipelines
 - Milvus vector database
 - GPU nodes (for embedding generation)

pipelines/pyproject.toml

@@ -0,0 +1,16 @@
+[project]
+name = "docs-agent-pipelines"
+version = "0.1.0"
+description = "Kubeflow Pipelines for RAG document indexing and embedding"
+requires-python = ">=3.11"
+dependencies = [
+    "kfp",
+    "requests",
+    "beautifulsoup4",
+    "sentence-transformers",
+    "langchain-text-splitters",
+    "torch",
+    "feast[milvus]",
+    "pandas",
+    "numpy",
+]

pyproject.toml

@@ -0,0 +1,25 @@
+[project]
+name = "docs-agent"
+version = "0.1.0"
+description = "Kubeflow Documentation AI Agent to power the Kubeflow Website"
+readme = "README.md"
+requires-python = ">=3.11"
+dependencies = []
+
+[tool.uv.workspace]
+members = [
+    "server",
+    "server-https",
+    "pipelines",
+]
+
+# PyTorch CPU-only index — avoids downloading the ~2 GB CUDA wheel
+[[tool.uv.index]]
+name = "pytorch-cpu"
+url = "https://download.pytorch.org/whl/cpu"
+explicit = true
+
+[tool.uv.sources]
+torch = { index = "pytorch-cpu" }
+torchvision = { index = "pytorch-cpu" }
+torchaudio = { index = "pytorch-cpu" }

server-https/Dockerfile

@@ -1,12 +1,16 @@
-    FROM python:3.11-slim
+FROM python:3.11-slim
+
+COPY --from=ghcr.io/astral-sh/uv:latest /uv /usr/local/bin/uv
 
 # Create non-root user
 RUN useradd -m -u 1000 appuser
 WORKDIR /app
 
-# Install deps
-COPY requirements.txt ./
-RUN pip install --no-cache-dir -r requirements.txt
+# Copy dependency definition and install with uv
+COPY pyproject.toml ./
+RUN uv pip install --system --no-cache \
+    --extra-index-url https://download.pytorch.org/whl/cpu \
+    -r pyproject.toml
 
 # Environment variables
 ENV PORT=8000

server-https/pyproject.toml

@@ -0,0 +1,15 @@
+[project]
+name = "docs-agent-server-https"
+version = "0.1.0"
+description = "FastAPI HTTPS server for the Kubeflow Documentation Agent"
+requires-python = ">=3.11"
+dependencies = [
+    "fastapi",
+    "uvicorn[standard]",
+    "pydantic",
+    "httpx",
+    "pymilvus",
+    "sentence-transformers",
+    "torch",
+    "numpy",
+]

server/Dockerfile

@@ -1,12 +1,16 @@
 FROM python:3.11-slim
 
+COPY --from=ghcr.io/astral-sh/uv:latest /uv /usr/local/bin/uv
+
 # Create non-root user
 RUN useradd -m -u 1000 appuser
 WORKDIR /app
 
-# Install deps
-COPY requirements.txt ./
-RUN pip install --no-cache-dir -r requirements.txt
+# Copy dependency definition and install with uv
+COPY pyproject.toml ./
+RUN uv pip install --system --no-cache \
+    --extra-index-url https://download.pytorch.org/whl/cpu \
+    -r pyproject.toml
 
 # Writable caches for HF/Transformers/Sentence-Transformers
 ENV HF_HOME=/home/appuser/.cache/huggingface \

server/pyproject.toml

@@ -0,0 +1,13 @@
+[project]
+name = "docs-agent-server"
+version = "0.1.0"
+description = "WebSocket server for the Kubeflow Documentation Agent"
+requires-python = ">=3.11"
+dependencies = [
+    "websockets",
+    "httpx",
+    "pymilvus",
+    "sentence-transformers",
+    "torch",
+    "numpy",
+]