Deploy to HuggingFace Spaces

.env.example

@@ -0,0 +1,27 @@
+# ==============================================
+# BACKEND CONFIGURATION (Hugging Face / Local)
+# ==============================================
+
+# Qdrant Cloud Credentials
+# Get these from https://cloud.qdrant.io
+# If running locally with Docker, use: http://localhost:6333 and leave API_KEY empty
+QDRANT_URL=https://your-cluster-id.region.qdrant.tech
+QDRANT_API_KEY=your-super-secret-api-key-here
+
+# Qdrant Settings (Optional)
+# QDRANT_COLLECTION=bio_discovery
+
+# ==============================================
+# FRONTEND CONFIGURATION (Vercel / Local)
+# ==============================================
+# NOTE: For Next.js (Vercel), these usually go in ui/.env.local or Vercel Dashboard
+
+# The URL where your Backend is running
+# Local: http://localhost:8000
+# Production (Hugging Face): https://your-space-name.hf.space
+NEXT_PUBLIC_API_URL=http://localhost:8000
+
+# The URL where your Frontend is running
+# Local: http://localhost:3000
+# Production (Vercel): https://your-project.vercel.app
+NEXT_PUBLIC_APP_URL=http://localhost:3000

Dockerfile.hf

@@ -0,0 +1,44 @@
+# Use a lightweight python image
+FROM python:3.9-slim
+
+# Set environment variables
+ENV PYTHONDONTWRITEBYTECODE=1 \
+    PYTHONUNBUFFERED=1 \
+    DEBIAN_FRONTEND=noninteractive
+
+# Install system dependencies required for RDKit and build tools
+RUN apt-get update && apt-get install -y \
+    libxrender1 \
+    libxext6 \
+    build-essential \
+    wget \
+    && rm -rf /var/lib/apt/lists/*
+
+# Set working directory
+WORKDIR /app
+
+# Upgrade pip
+RUN pip install --no-cache-dir --upgrade pip
+
+# Copy only requirements first to cache dependencies
+COPY requirements.txt .
+
+# Install dependencies
+# Using automatic cpu version for torch to keep image small
+RUN pip install --no-cache-dir torch torchvision --index-url https://download.pytorch.org/whl/cpu
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy the rest of the application
+COPY . .
+
+# Expose the API port
+EXPOSE 8000
+
+# Create a non-root user for security (good practice for HF Spaces)
+RUN useradd -m -u 1000 user
+USER user
+ENV HOME=/home/user \
+    PATH=/home/user/.local/bin:$PATH
+
+# Command to run the application
+CMD ["uvicorn", "bioflow.api.server:app", "--host", "0.0.0.0", "--port", "8000"]

bioflow/api/qdrant_service.py

@@ -88,6 +88,7 @@ class QdrantService:
         self,
         model_service=None,
         url: str = None,
+        api_key: str = None,
         path: str = None,
         vector_dim: int = 768
     ):
@@ -97,6 +98,7 @@ class QdrantService:
         Args:
             model_service: ModelService for embeddings
             url: Qdrant server URL (e.g., http://localhost:6333)
+            api_key: Qdrant API key for cloud clusters
             path: Path for local Qdrant storage
             vector_dim: Dimension of embedding vectors
             
@@ -105,6 +107,7 @@ class QdrantService:
         """
         self.model_service = model_service
         self.url = url or os.getenv("QDRANT_URL")
+        self.api_key = api_key or os.getenv("QDRANT_API_KEY")
         self.path = path or os.getenv("QDRANT_PATH", "./qdrant_data")
         self.vector_dim = vector_dim
         self.hnsw_m = int(os.getenv("QDRANT_HNSW_M", "16"))
@@ -143,7 +146,7 @@ class QdrantService:
             from qdrant_client import QdrantClient
             
             if self.url:
-                self._client = QdrantClient(url=self.url)
+                self._client = QdrantClient(url=self.url, api_key=self.api_key)
                 logger.info(f"Connected to Qdrant at {self.url}")
             else:
                 self._client = QdrantClient(path=self.path)

requirements.txt

@@ -20,3 +20,5 @@ scanpy==1.10.3
 qdrant-client>=1.7.0
 plotly>=5.18.0
 scikit-learn>=1.3.0
+
+torch>=2.0.0

server/api.py

@@ -36,9 +36,10 @@ from config import (
 app = FastAPI(title="BioDiscovery API", version="2.0")
 
 # CORS for frontend
+# Allow generic access for deployment - in production restrict this to your Vercel domain
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["http://localhost:3000", "http://127.0.0.1:3000"],
+    allow_origins=["*"], 
     allow_methods=["*"],
     allow_headers=["*"],
 )