Add Docker support and remove Ollama

DOCKER_COMMANDS.md

@@ -0,0 +1,266 @@
+# Docker Commands for IntegraChat
+
+## Quick Start Commands
+
+### 1. Stop and Remove Existing Container
+```powershell
+docker rm -f integrachat
+```
+
+### 2. Build the Docker Image
+```powershell
+docker build -t integrachat:latest .
+```
+
+### 3. Run the Container
+```powershell
+docker run -d --name integrachat `
+    -p 7860:7860 `
+    -p 8000:8000 `
+    -p 8900:8900 `
+    --env-file .env `
+    -e DOCKER_CONTAINER=1 `
+    integrachat:latest
+```
+
+### 4. View Logs
+```powershell
+# Follow logs (live)
+docker logs -f integrachat
+
+# Last 50 lines
+docker logs --tail 50 integrachat
+
+# Last 100 lines with timestamps
+docker logs --tail 100 -t integrachat
+```
+
+---
+
+## Container Management
+
+### Check Container Status
+```powershell
+# Check if running
+docker ps --filter "name=integrachat"
+
+# Check all containers (including stopped)
+docker ps -a --filter "name=integrachat"
+
+# Detailed status
+docker ps --filter "name=integrachat" --format "Container: {{.Names}} | Status: {{.Status}} | Ports: {{.Ports}}"
+```
+
+### Stop Container
+```powershell
+docker stop integrachat
+```
+
+### Start Container (if stopped)
+```powershell
+docker start integrachat
+```
+
+### Restart Container
+```powershell
+docker restart integrachat
+```
+
+### Remove Container
+```powershell
+# Stop and remove
+docker rm -f integrachat
+
+# Remove only if stopped
+docker rm integrachat
+```
+
+---
+
+## Image Management
+
+### List Images
+```powershell
+docker images integrachat
+```
+
+### Remove Image
+```powershell
+docker rmi integrachat:latest
+```
+
+### Rebuild Without Cache
+```powershell
+docker build --no-cache -t integrachat:latest .
+```
+
+---
+
+## Debugging Commands
+
+### Execute Commands Inside Container
+```powershell
+# Open shell in container
+docker exec -it integrachat /bin/bash
+
+# Run Python command
+docker exec integrachat python --version
+
+# Check environment variables
+docker exec integrachat printenv | Select-String -Pattern "GROQ|MCP|API"
+
+# Check if services are running
+docker exec integrachat ps aux
+```
+
+### Check Service Health
+```powershell
+# Check FastAPI health
+docker exec integrachat curl -s http://localhost:8000/health
+
+# Check MCP server health
+docker exec integrachat curl -s http://localhost:8900/health
+
+# Check Gradio (from host)
+Invoke-WebRequest -Uri http://localhost:7860 -UseBasicParsing -TimeoutSec 5
+```
+
+### View Service Logs
+```powershell
+# FastAPI logs
+docker exec integrachat tail -n 50 /app/logs/fastapi.log
+
+# MCP server logs
+docker exec integrachat tail -n 50 /app/logs/mcp.log
+
+# Gradio logs
+docker exec integrachat tail -n 50 /app/logs/gradio.log
+
+# All logs
+docker exec integrachat tail -n 50 /app/logs/*.log
+```
+
+### Check Ports
+```powershell
+# Check what ports are mapped
+docker port integrachat
+
+# Check if ports are listening (from host)
+netstat -an | Select-String -Pattern "7860|8000|8900"
+```
+
+---
+
+## Complete Rebuild Sequence
+
+```powershell
+# 1. Stop and remove container
+docker rm -f integrachat
+
+# 2. Remove old image (optional)
+docker rmi integrachat:latest
+
+# 3. Build new image
+docker build -t integrachat:latest .
+
+# 4. Run container
+docker run -d --name integrachat `
+    -p 7860:7860 `
+    -p 8000:8000 `
+    -p 8900:8900 `
+    --env-file .env `
+    -e DOCKER_CONTAINER=1 `
+    integrachat:latest
+
+# 5. Check status
+docker ps --filter "name=integrachat"
+
+# 6. View logs
+docker logs -f integrachat
+```
+
+---
+
+## Quick Health Check
+
+```powershell
+# Check all services
+Write-Host "Container Status:" -ForegroundColor Cyan
+docker ps --filter "name=integrachat" --format "  {{.Names}}: {{.Status}}"
+
+Write-Host "`nService Health:" -ForegroundColor Cyan
+Write-Host "  FastAPI:" -NoNewline
+docker exec integrachat curl -s http://localhost:8000/health 2>&1 | Out-Null
+if ($LASTEXITCODE -eq 0) { Write-Host " ✓ Running" -ForegroundColor Green } else { Write-Host " ✗ Not responding" -ForegroundColor Red }
+
+Write-Host "  MCP Server:" -NoNewline
+docker exec integrachat curl -s http://localhost:8900/health 2>&1 | Out-Null
+if ($LASTEXITCODE -eq 0) { Write-Host " ✓ Running" -ForegroundColor Green } else { Write-Host " ✗ Not responding" -ForegroundColor Red }
+
+Write-Host "  Gradio UI:" -NoNewline
+try { $response = Invoke-WebRequest -Uri http://localhost:7860 -UseBasicParsing -TimeoutSec 2; Write-Host " ✓ Running" -ForegroundColor Green } catch { Write-Host " ✗ Not responding" -ForegroundColor Red }
+```
+
+---
+
+## Access URLs
+
+Once the container is running, access:
+
+- **Gradio UI**: http://localhost:7860
+- **FastAPI API**: http://localhost:8000
+- **API Docs**: http://localhost:8000/docs
+- **MCP Server**: http://localhost:8900
+- **MCP Server Docs**: http://localhost:8900/docs
+
+---
+
+## Troubleshooting
+
+### Container won't start
+```powershell
+# Check logs for errors
+docker logs integrachat
+
+# Check if ports are already in use
+netstat -an | Select-String -Pattern "7860|8000|8900"
+```
+
+### Services not responding
+```powershell
+# Restart container
+docker restart integrachat
+
+# Check service logs inside container
+docker exec integrachat tail -n 100 /app/logs/*.log
+```
+
+### Clear everything and start fresh
+```powershell
+# Stop and remove container
+docker rm -f integrachat
+
+# Remove image
+docker rmi integrachat:latest
+
+# Clear build cache (optional)
+docker builder prune -f
+
+# Rebuild from scratch
+docker build --no-cache -t integrachat:latest .
+```
+
+---
+
+## Environment Variables
+
+Make sure your `.env` file has:
+- `GROQ_API_KEY` - Your Groq API key
+- `GROQ_MODEL` - Model name (default: llama-3.1-8b-instant)
+- `RAG_MCP_URL` - http://localhost:8900/rag
+- `WEB_MCP_URL` - http://localhost:8900/web
+- `ADMIN_MCP_URL` - http://localhost:8900/admin
+- `MCP_PORT` - 8900
+- `API_PORT` - 8000
+- `POSTGRESQL_URL` - Your database connection string (optional)
+

DOCKER_GUIDE.md

@@ -0,0 +1,225 @@
+# Docker Setup Guide for IntegraChat
+
+## Quick Start
+
+### Option 1: Use PowerShell Script (Easiest for Windows)
+
+```powershell
+# Run the helper script
+.\run-docker.ps1
+```
+
+### Option 2: Build and Run with Docker Compose (Recommended)
+
+```powershell
+# PowerShell
+docker-compose up -d
+
+# Or with rebuild
+docker-compose up -d --build
+```
+
+### Build and Run Manually
+
+**PowerShell (Windows):**
+```powershell
+# Build the image
+docker build -t integrachat:latest .
+
+# Run the container (PowerShell uses backticks for line continuation)
+docker run -d --name integrachat `
+  -p 7860:7860 -p 8000:8000 -p 8900:8900 `
+  -e DOCKER_CONTAINER=1 `
+  integrachat:latest
+
+# Or use a single line:
+docker run -d --name integrachat -p 7860:7860 -p 8000:8000 -p 8900:8900 -e DOCKER_CONTAINER=1 integrachat:latest
+```
+
+**Bash/Linux/Mac:**
+```bash
+# Build the image
+docker build -t integrachat:latest .
+
+# Run the container
+docker run -d --name integrachat \
+  -p 7860:7860 -p 8000:8000 -p 8900:8900 \
+  -e DOCKER_CONTAINER=1 \
+  integrachat:latest
+```
+
+## Container Management
+
+### View Logs
+```bash
+# All logs (streaming)
+docker logs -f integrachat
+
+# Specific service logs
+docker exec integrachat tail -f /app/logs/fastapi.log
+docker exec integrachat tail -f /app/logs/gradio.log
+docker exec integrachat tail -f /app/logs/mcp.log
+```
+
+### Stop Container
+```bash
+docker stop integrachat
+```
+
+### Start Container
+```bash
+docker start integrachat
+```
+
+### Remove Container
+```bash
+docker stop integrachat
+docker rm integrachat
+```
+
+### Rebuild After Changes
+
+**PowerShell:**
+```powershell
+docker stop integrachat
+docker rm integrachat
+docker build -t integrachat:latest .
+docker run -d --name integrachat -p 7860:7860 -p 8000:8000 -p 8900:8900 -e DOCKER_CONTAINER=1 integrachat:latest
+```
+
+**Bash:**
+```bash
+docker stop integrachat
+docker rm integrachat
+docker build -t integrachat:latest .
+docker run -d --name integrachat \
+  -p 7860:7860 -p 8000:8000 -p 8900:8900 \
+  -e DOCKER_CONTAINER=1 \
+  integrachat:latest
+```
+
+## Access Services
+
+- **Gradio UI**: http://localhost:7860
+- **FastAPI API**: http://localhost:8000
+- **MCP Server**: http://localhost:8900
+- **API Docs**: http://localhost:8000/docs
+- **MCP Docs**: http://localhost:8900/docs
+
+## Environment Variables
+
+Create a `.env` file (or use docker-compose.yml) to configure:
+
+```env
+# LLM Configuration
+LLM_BACKEND=groq  # or "ollama"
+GROQ_API_KEY=your_key_here
+GROQ_MODEL=llama-3.1-8b-instant
+
+# Supabase (optional - for analytics)
+SUPABASE_URL=https://your-project.supabase.co
+SUPABASE_SERVICE_KEY=your_service_key
+
+# Ports (defaults shown)
+API_PORT=8000
+MCP_PORT=8900
+GRADIO_PORT=7860
+```
+
+## Docker Compose
+
+The `docker-compose.yml` file provides:
+- Easy service management
+- Environment variable support
+- Volume mounting for logs
+- Health checks
+- Auto-restart on failure
+
+### Using Docker Compose
+
+```bash
+# Start services
+docker-compose up -d
+
+# View logs
+docker-compose logs -f
+
+# Stop services
+docker-compose down
+
+# Rebuild and restart
+docker-compose up -d --build
+```
+
+## PowerShell-Specific Notes
+
+### Line Continuation
+PowerShell uses backticks (`` ` ``) for line continuation, not backslashes (`\`):
+
+```powershell
+# ✅ Correct (PowerShell)
+docker run -d --name integrachat `
+  -p 7860:7860 `
+  -p 8000:8000 `
+  integrachat:latest
+
+# ✅ Also correct (single line)
+docker run -d --name integrachat -p 7860:7860 -p 8000:8000 -p 8900:8900 -e DOCKER_CONTAINER=1 integrachat:latest
+
+# ❌ Wrong (bash syntax - doesn't work in PowerShell)
+docker run -d --name integrachat \
+  -p 7860:7860 \
+  integrachat:latest
+```
+
+### Quick Commands Script
+Use `run-docker.ps1` for easy container management:
+```powershell
+.\run-docker.ps1
+```
+
+## Troubleshooting
+
+### Check Container Status
+```bash
+docker ps -a | grep integrachat
+```
+
+### Check Service Health
+```bash
+# FastAPI health
+curl http://localhost:8000/health
+
+# MCP health
+curl http://localhost:8900/health
+```
+
+### View All Logs
+```bash
+docker exec integrachat tail -n 100 /app/logs/*.log
+```
+
+### Restart Services Inside Container
+```bash
+# Container will auto-restart services, but you can manually restart:
+docker restart integrachat
+```
+
+### Clean Up
+```bash
+# Remove container and image
+docker stop integrachat
+docker rm integrachat
+docker rmi integrachat:latest
+
+# Remove all unused Docker resources
+docker system prune -a
+```
+
+## Notes
+
+- The container runs all three services (MCP, FastAPI, Gradio) automatically
+- Logs are written to `/app/logs/` inside the container
+- The entrypoint script handles service startup and health checks
+- Supabase warnings are expected if credentials are not configured (analytics will be disabled gracefully)
+

Dockerfile

@@ -8,6 +8,7 @@ WORKDIR /app
 # Install system dependencies
 RUN apt-get update && \
     apt-get install -y --no-install-recommends \
+        --fix-missing \
         build-essential \
         curl \
         git && \

backend/api/mcp_clients/rag_client.py

@@ -11,7 +11,7 @@ class RAGClient:
     """
 
     def __init__(self):
-        self.base_url = os.getenv("RAG_MCP_URL", "http://localhost:8001")
+        self.base_url = os.getenv("RAG_MCP_URL", "http://localhost:8900/rag")
         if not self.base_url:
             raise ValueError("RAG_MCP_URL environment variable is not set")
         self.search_endpoint = f"{self.base_url}/search"

backend/api/routes/admin.py

@@ -39,11 +39,16 @@ def _get_analytics_store() -> Optional[AnalyticsStore]:
     try:
         _analytics_store = AnalyticsStore()
     except RuntimeError as exc:
-        logger.warning("Admin analytics disabled: %s", exc)
+        # Only log at warning level if credentials are configured (actual error)
+        # Otherwise log at debug level (expected when Supabase is not configured)
+        if os.getenv("SUPABASE_URL") and os.getenv("SUPABASE_SERVICE_KEY"):
+            logger.warning("Analytics disabled: %s", str(exc).split('\n')[0])  # Only first line
+        else:
+            logger.debug("Analytics disabled: %s", str(exc).split('\n')[0])
         _analytics_failed = True
         _analytics_store = None
     except Exception as exc:  # pragma: no cover - unexpected failures
-        logger.debug("Admin analytics unexpected init failure: %s", exc)
+        logger.debug("Analytics unexpected init failure: %s", exc)
         _analytics_failed = True
         _analytics_store = None
 

backend/api/routes/agent.py

@@ -23,10 +23,9 @@ router = APIRouter()
 
 
 orchestrator = AgentOrchestrator(
-    rag_mcp_url=os.getenv("RAG_MCP_URL", "http://localhost:8001"),
-    web_mcp_url=os.getenv("WEB_MCP_URL", "http://localhost:8002"),
-    admin_mcp_url=os.getenv("ADMIN_MCP_URL", "http://localhost:8003"),
-    llm_backend=os.getenv("LLM_BACKEND", "ollama")
+    rag_mcp_url=os.getenv("RAG_MCP_URL", "http://localhost:8900/rag"),
+    web_mcp_url=os.getenv("WEB_MCP_URL", "http://localhost:8900/web"),
+    admin_mcp_url=os.getenv("ADMIN_MCP_URL", "http://localhost:8900/admin")
 )
 
 

backend/api/routes/analytics.py

@@ -19,17 +19,24 @@ try:
         analytics_store: Optional[AnalyticsStore] = AnalyticsStore()
     else:
         analytics_store = None
-        logger.warning(
+        logger.debug(
             "AnalyticsStore: Supabase credentials not configured. "
             "Analytics endpoints will return 503."
         )
 except RuntimeError as exc:
     analytics_store = None
-    logger.warning(
-        "AnalyticsStore initialization failed (%s). "
-        "Analytics endpoints will return 503.",
-        exc,
-    )
+    # Only log at warning level if credentials are configured (actual error)
+    # Otherwise log at debug level (expected when Supabase is not configured)
+    if os.getenv("SUPABASE_URL") and os.getenv("SUPABASE_SERVICE_KEY"):
+        logger.warning(
+            "AnalyticsStore initialization failed (%s). Analytics endpoints will return 503.",
+            str(exc).split('\n')[0],  # Only first line
+        )
+    else:
+        logger.debug(
+            "AnalyticsStore not configured (%s). Analytics endpoints will return 503.",
+            str(exc).split('\n')[0],
+        )
 
 
 @router.get("/overview")
@@ -43,16 +50,30 @@ async def analytics_overview(
     Includes total queries, tool usage, red-flag count, and active users.
     """
 
-    if analytics_store is None:
-        raise HTTPException(
-            status_code=503,
-            detail="Analytics is disabled because Supabase is not configured on this deployment.",
-        )
-
     if not x_tenant_id:
         raise HTTPException(status_code=400, detail="Missing tenant ID")
     require_api_permission(x_user_role, "view_analytics")
 
+    # Return empty data if analytics is not configured (instead of 503)
+    if analytics_store is None:
+        return {
+            "tenant_id": x_tenant_id,
+            "overview": {
+                "total_queries": 0,
+                "tool_usage": {},
+                "redflag_count": 0,
+                "active_users": 0,
+                "last_query": None,
+                "rag_quality": {
+                    "total_searches": 0,
+                    "avg_hits_per_search": 0,
+                    "avg_score": 0.0,
+                    "avg_top_score": 0.0,
+                    "avg_latency_ms": 0.0
+                }
+            }
+        }
+
     since_timestamp = int((datetime.now() - timedelta(days=days)).timestamp()) if days else None
     
     tool_usage = analytics_store.get_tool_usage_stats(x_tenant_id, since_timestamp)
@@ -83,16 +104,18 @@ async def analytics_tool_usage(
     Includes counts, latency, tokens, and success/error rates.
     """
 
-    if analytics_store is None:
-        raise HTTPException(
-            status_code=503,
-            detail="Analytics is disabled because Supabase is not configured on this deployment.",
-        )
-
     if not x_tenant_id:
         raise HTTPException(status_code=400, detail="Missing tenant ID")
     require_api_permission(x_user_role, "view_analytics")
 
+    # Return empty data if analytics is not configured (instead of 503)
+    if analytics_store is None:
+        return {
+            "tenant_id": x_tenant_id,
+            "tool_usage": {},
+            "period_days": days
+        }
+
     since_timestamp = int((datetime.now() - timedelta(days=days)).timestamp()) if days else None
     tool_usage = analytics_store.get_tool_usage_stats(x_tenant_id, since_timestamp)
 
@@ -115,16 +138,18 @@ async def analytics_redflags(
     Includes rule details, severity, confidence, and timestamps.
     """
 
-    if analytics_store is None:
-        raise HTTPException(
-            status_code=503,
-            detail="Analytics is disabled because Supabase is not configured on this deployment.",
-        )
-
     if not x_tenant_id:
         raise HTTPException(status_code=400, detail="Missing tenant ID")
     require_api_permission(x_user_role, "view_analytics")
 
+    # Return empty data if analytics is not configured (instead of 503)
+    if analytics_store is None:
+        return {
+            "tenant_id": x_tenant_id,
+            "redflags": [],
+            "count": 0
+        }
+
     since_timestamp = int((datetime.now() - timedelta(days=days)).timestamp()) if days else None
     redflags = analytics_store.get_redflag_violations(x_tenant_id, limit, since_timestamp)
 
@@ -151,16 +176,24 @@ async def analytics_activity(
     Includes total queries, active users, last query timestamp, and individual activity records for heatmap visualization.
     """
 
-    if analytics_store is None:
-        raise HTTPException(
-            status_code=503,
-            detail="Analytics is disabled because Supabase is not configured on this deployment.",
-        )
-
     if not x_tenant_id:
         raise HTTPException(status_code=400, detail="Missing tenant ID")
     require_api_permission(x_user_role, "view_analytics")
 
+    # Return empty data if analytics is not configured (instead of 503)
+    if analytics_store is None:
+        return {
+            "tenant_id": x_tenant_id,
+            "activity": {
+                "total_queries": 0,
+                "active_users": 0,
+                "redflag_count": 0,
+                "last_query": None
+            },
+            "activities": [],
+            "period_days": days
+        }
+
     since_timestamp = int((datetime.now() - timedelta(days=days)).timestamp()) if days else None
     activity = analytics_store.get_activity_summary(x_tenant_id, since_timestamp)
     
@@ -186,16 +219,24 @@ async def analytics_rag_quality(
     Includes average hits, scores, and latency.
     """
 
-    if analytics_store is None:
-        raise HTTPException(
-            status_code=503,
-            detail="Analytics is disabled because Supabase is not configured on this deployment.",
-        )
-
     if not x_tenant_id:
         raise HTTPException(status_code=400, detail="Missing tenant ID")
     require_api_permission(x_user_role, "view_analytics")
 
+    # Return empty data if analytics is not configured (instead of 503)
+    if analytics_store is None:
+        return {
+            "tenant_id": x_tenant_id,
+            "rag_quality": {
+                "total_searches": 0,
+                "avg_hits_per_search": 0,
+                "avg_score": 0.0,
+                "avg_top_score": 0.0,
+                "avg_latency_ms": 0.0
+            },
+            "period_days": days
+        }
+
     since_timestamp = int((datetime.now() - timedelta(days=days)).timestamp()) if days else None
     rag_quality = analytics_store.get_rag_quality_metrics(x_tenant_id, since_timestamp)
 

backend/api/routes/rag.py

@@ -261,7 +261,7 @@ async def rag_ingest_document(
             error_msg = (
                 f"RAG server error: {error_detail}\n\n"
                 f"Please check:\n"
-                f"1. RAG_MCP_URL is set correctly (default: http://localhost:8001)\n"
+                f"1. RAG_MCP_URL is set correctly (default: http://localhost:8900/rag)\n"
                 f"2. RAG MCP server is running\n"
                 f"3. Database connection (POSTGRESQL_URL) is configured in the RAG server"
             )

backend/api/services/agent_orchestrator.py

@@ -40,9 +40,10 @@ load_dotenv()
 
 class AgentOrchestrator:
 
-    def __init__(self, rag_mcp_url: str, web_mcp_url: str, admin_mcp_url: str, llm_backend: str = "ollama"):
+    def __init__(self, rag_mcp_url: str, web_mcp_url: str, admin_mcp_url: str):
         self.mcp = MCPClient(rag_mcp_url, web_mcp_url, admin_mcp_url)
-        self.llm = LLMClient(backend=llm_backend, url=os.getenv("OLLAMA_URL"), api_key=os.getenv("GROQ_API_KEY"), model=os.getenv("OLLAMA_MODEL"))
+        # Groq-only LLM client
+        self.llm = LLMClient(api_key=os.getenv("GROQ_API_KEY"), model=os.getenv("GROQ_MODEL"))
         
         # pass admin_mcp_url so detector can call back
         self.redflag = RedFlagDetector(
@@ -68,15 +69,20 @@ class AgentOrchestrator:
             return
 
         if self._analytics_disabled:
-            print("⚠️  AgentOrchestrator Analytics: Disabled via ANALYTICS_DISABLED")
+            logger.info("Analytics: Disabled via ANALYTICS_DISABLED")
         else:
             store = self._get_analytics()
             if store is None:
-                print("⚠️  AgentOrchestrator Analytics: Disabled (Supabase not configured)")
+                # Only log if credentials might be missing (not if package is missing)
+                import os
+                if os.getenv("SUPABASE_URL") and os.getenv("SUPABASE_SERVICE_KEY"):
+                    logger.warning("Analytics: Disabled (Supabase initialization failed)")
+                else:
+                    logger.debug("Analytics: Disabled (Supabase not configured)")
             elif store.use_supabase:
-                print("✅ AgentOrchestrator Analytics: Using Supabase backend")
+                logger.info("Analytics: Using Supabase backend")
             else:
-                print("⚠️  AgentOrchestrator Analytics: Using fallback backend")
+                logger.warning("Analytics: Using fallback backend")
 
         AgentOrchestrator._analytics_backend_logged = True
 
@@ -90,11 +96,17 @@ class AgentOrchestrator:
         try:
             self._analytics = AnalyticsStore()
         except RuntimeError as exc:
-            logger.warning("AgentOrchestrator analytics disabled: %s", exc)
+            # Only log at warning level if credentials are configured (actual error)
+            # Otherwise log at debug level (expected when Supabase is not configured)
+            import os
+            if os.getenv("SUPABASE_URL") and os.getenv("SUPABASE_SERVICE_KEY"):
+                logger.warning("Analytics disabled: %s", str(exc).split('\n')[0])  # Only first line
+            else:
+                logger.debug("Analytics disabled: %s", str(exc).split('\n')[0])
             self._analytics_failed = True
             self._analytics = None
         except Exception as exc:  # pragma: no cover - unexpected initialization failures
-            logger.debug("AgentOrchestrator analytics unexpected init failure: %s", exc)
+            logger.debug("Analytics unexpected init failure: %s", exc)
             self._analytics_failed = True
             self._analytics = None
 
@@ -1169,14 +1181,13 @@ Answer:"""
                     fallback = await self.llm.simple_call(req.message, temperature=req.temperature)
                 except Exception as llm_error:
                     error_msg = str(llm_error)
-                    if "Cannot connect" in error_msg or "Ollama" in error_msg:
+                    if "Groq API key" in error_msg or "GROQ_API_KEY" in error_msg:
                         fallback = (
                             f"I encountered an error while processing your request: {str(e)}\n\n"
-                            f"Additionally, the AI service (Ollama) is unavailable: {error_msg}\n\n"
+                            f"Additionally, the AI service (Groq) is unavailable: {error_msg}\n\n"
                             f"To fix:\n"
-                            f"1. Install Ollama from https://ollama.ai\n"
-                            f"2. Start: `ollama serve`\n"
-                            f"3. Pull model: `ollama pull {os.getenv('OLLAMA_MODEL', 'llama3.1:latest')}`"
+                            f"1. Get a free Groq API key from https://console.groq.com\n"
+                            f"2. Set GROQ_API_KEY in your .env file or environment variables"
                         )
                     else:
                         fallback = f"I encountered an error while processing your request: {str(e)}. Additionally, the AI service is unavailable: {error_msg}"
@@ -1315,15 +1326,14 @@ Answer:"""
         except Exception as e:
             # If LLM fails, return a helpful error message
             error_msg = str(e)
-            if "Cannot connect" in error_msg or "Ollama" in error_msg:
+            if "Groq API key" in error_msg or "GROQ_API_KEY" in error_msg:
                 llm_out = (
-                    f"I couldn't connect to the AI service (Ollama). "
+                    f"I couldn't connect to the AI service (Groq). "
                     f"Error: {error_msg}\n\n"
                     f"To fix this:\n"
-                    f"1. Install Ollama from https://ollama.ai\n"
-                    f"2. Start Ollama: `ollama serve`\n"
-                    f"3. Pull the model: `ollama pull {os.getenv('OLLAMA_MODEL', 'llama3.1:latest')}`\n"
-                    f"4. Or set OLLAMA_URL and OLLAMA_MODEL in your .env file"
+                    f"1. Get a free Groq API key from https://console.groq.com\n"
+                    f"2. Set GROQ_API_KEY in your .env file or environment variables\n"
+                    f"3. Optionally set GROQ_MODEL (default: llama-3.1-8b-instant)"
                 )
             else:
                 llm_out = f"I apologize, but I'm unable to process your request right now. The AI service is unavailable: {error_msg}"
@@ -1997,15 +2007,14 @@ Answer:"""
             tool_traces.append({"tool": "llm", "error": str(e)})
             error_msg = str(e)
             # Provide helpful error message
-            if "Cannot connect" in error_msg or "Ollama" in error_msg:
+            if "Groq API key" in error_msg or "GROQ_API_KEY" in error_msg:
                 fallback = (
-                    f"I couldn't connect to the AI service (Ollama). "
+                    f"I couldn't connect to the AI service (Groq). "
                     f"Error: {error_msg}\n\n"
                     f"To fix this:\n"
-                    f"1. Install Ollama from https://ollama.ai\n"
-                    f"2. Start Ollama: `ollama serve`\n"
-                    f"3. Pull the model: `ollama pull {os.getenv('OLLAMA_MODEL', 'llama3.1:latest')}`\n"
-                    f"4. Or set OLLAMA_URL and OLLAMA_MODEL in your .env file"
+                    f"1. Get a free Groq API key from https://console.groq.com\n"
+                    f"2. Set GROQ_API_KEY in your .env file or environment variables\n"
+                    f"3. Optionally set GROQ_MODEL (default: llama-3.1-8b-instant)"
                 )
             else:
                 fallback = f"I encountered an error while synthesizing the response: {error_msg}"

backend/api/services/document_ingestion.py

@@ -324,7 +324,7 @@ async def process_ingestion(
         raise RuntimeError(
             f"Failed to send document to RAG MCP server: {str(e)}\n\n"
             f"Please check:\n"
-            f"1. RAG_MCP_URL is set correctly (default: http://localhost:8001)\n"
+            f"1. RAG_MCP_URL is set correctly (default: http://localhost:8900/rag)\n"
             f"2. RAG MCP server is running\n"
             f"3. Database connection (POSTGRESQL_URL) is configured in the RAG server"
         ) from e

backend/api/services/llm_client.py

@@ -5,68 +5,65 @@ from typing import AsyncGenerator
 
 class LLMClient:
 
-    def __init__(self, backend="ollama", url=None, api_key=None, model=None):
-        self.backend = backend
-        self.url = url or os.getenv("OLLAMA_URL", "http://localhost:11434")
+    def __init__(self, api_key=None, model=None):
         self.api_key = api_key or os.getenv("GROQ_API_KEY")
-        # Default model based on backend
-        if backend == "groq":
-            self.model = model or os.getenv("GROQ_MODEL", "llama-3.1-70b-versatile")
-        else:
-            self.model = model or os.getenv("OLLAMA_MODEL", "llama3.1:latest")
+        self.model = model or os.getenv("GROQ_MODEL", "llama-3.1-8b-instant")
         self.http = httpx.AsyncClient(timeout=30)
 
-
     async def simple_call(self, prompt: str, temperature: float = 0.0) -> str:
-        if self.backend=="ollama":
-            if not self.url or not self.model:
-                raise RuntimeError(f"LLM not configured: url={self.url}, model={self.model}. Set OLLAMA_URL and OLLAMA_MODEL env vars.")
-            
-            try:
-                # Ollama uses /api/generate endpoint
-                r = await self.http.post(
-                    f"{self.url}/api/generate",
-                    json={
-                        "model": self.model,
-                        "prompt": prompt,
-                        "stream": False,
-                        "options": {"temperature": temperature}
-                    }
-                )
-                r.raise_for_status()
-                response_data = r.json()
-                return response_data.get("response", "")
-            except httpx.HTTPStatusError as e:
-                if e.response.status_code == 404:
-                    raise RuntimeError(
-                        f"Ollama endpoint not found. Is Ollama running at {self.url}? "
-                        f"Or does the model '{self.model}' exist? "
-                        f"Try: ollama pull {self.model}"
-                    )
-                elif e.response.status_code == 400:
-                    error_detail = e.response.json().get("error", "Unknown error")
-                    raise RuntimeError(f"Ollama API error: {error_detail}")
-                else:
-                    raise RuntimeError(f"Ollama API error: HTTP {e.response.status_code} - {e.response.text}")
-            except httpx.ConnectError:
-                raise RuntimeError(
-                    f"Cannot connect to Ollama at {self.url}. "
-                    f"Is Ollama running? Start it with: ollama serve"
-                )
-            except Exception as e:
-                raise RuntimeError(f"LLM call failed: {str(e)}")
-        elif self.backend == "groq":
-            if not self.api_key:
-                raise RuntimeError(
-                    "Groq API key not configured. Set GROQ_API_KEY environment variable. "
-                    "Get a free API key at https://console.groq.com"
-                )
-            if not self.model:
-                raise RuntimeError("Groq model not configured. Set GROQ_MODEL environment variable.")
-            
+        if not self.api_key:
+            raise RuntimeError(
+                "Groq API key not configured. Set GROQ_API_KEY environment variable. "
+                "Get a free API key at https://console.groq.com"
+            )
+        if not self.model:
+            raise RuntimeError("Groq model not configured. Set GROQ_MODEL environment variable.")
+        
+        try:
+            # Groq uses OpenAI-compatible API
+            r = await self.http.post(
+                "https://api.groq.com/openai/v1/chat/completions",
+                headers={
+                    "Authorization": f"Bearer {self.api_key}",
+                    "Content-Type": "application/json"
+                },
+                json={
+                    "model": self.model,
+                    "messages": [
+                        {"role": "user", "content": prompt}
+                    ],
+                    "temperature": temperature,
+                    "stream": False
+                }
+            )
+            r.raise_for_status()
+            response_data = r.json()
+            return response_data["choices"][0]["message"]["content"]
+        except httpx.HTTPStatusError as e:
+            error_detail = "Unknown error"
             try:
-                # Groq uses OpenAI-compatible API
-                r = await self.http.post(
+                error_json = e.response.json()
+                error_detail = error_json.get("error", {}).get("message", str(error_json))
+            except:
+                error_detail = e.response.text
+            raise RuntimeError(f"Groq API error: HTTP {e.response.status_code} - {error_detail}")
+        except Exception as e:
+            raise RuntimeError(f"Groq API call failed: {str(e)}")
+
+    async def stream_call(self, prompt: str, temperature: float = 0.0) -> AsyncGenerator[str, None]:
+        """Stream LLM response token by token."""
+        if not self.api_key:
+            raise RuntimeError(
+                "Groq API key not configured. Set GROQ_API_KEY environment variable. "
+                "Get a free API key at https://console.groq.com"
+            )
+        if not self.model:
+            raise RuntimeError("Groq model not configured. Set GROQ_MODEL environment variable.")
+        
+        try:
+            async with httpx.AsyncClient(timeout=300.0) as client:
+                async with client.stream(
+                    "POST",
                     "https://api.groq.com/openai/v1/chat/completions",
                     headers={
                         "Authorization": f"Bearer {self.api_key}",
@@ -78,117 +75,32 @@ class LLMClient:
                             {"role": "user", "content": prompt}
                         ],
                         "temperature": temperature,
-                        "stream": False
+                        "stream": True
                     }
-                )
-                r.raise_for_status()
-                response_data = r.json()
-                return response_data["choices"][0]["message"]["content"]
-            except httpx.HTTPStatusError as e:
-                error_detail = "Unknown error"
-                try:
-                    error_json = e.response.json()
-                    error_detail = error_json.get("error", {}).get("message", str(error_json))
-                except:
-                    error_detail = e.response.text
-                raise RuntimeError(f"Groq API error: HTTP {e.response.status_code} - {error_detail}")
-            except Exception as e:
-                raise RuntimeError(f"Groq API call failed: {str(e)}")
-        else:
-            raise RuntimeError(f"Unsupported backend: {self.backend}. Supported backends: 'ollama', 'groq'")
-
-    async def stream_call(self, prompt: str, temperature: float = 0.0) -> AsyncGenerator[str, None]:
-        """Stream LLM response token by token."""
-        if self.backend == "ollama":
-            if not self.url or not self.model:
-                raise RuntimeError(f"LLM not configured: url={self.url}, model={self.model}")
-            
-            try:
-                async with httpx.AsyncClient(timeout=300.0) as client:
-                    async with client.stream(
-                        "POST",
-                        f"{self.url}/api/generate",
-                        json={
-                            "model": self.model,
-                            "prompt": prompt,
-                            "stream": True,
-                            "options": {"temperature": temperature}
-                        }
-                    ) as response:
-                        response.raise_for_status()
-                        async for line in response.aiter_lines():
-                            if line:
+                ) as response:
+                    response.raise_for_status()
+                    async for line in response.aiter_lines():
+                        if line:
+                            # Groq uses Server-Sent Events format
+                            if line.startswith("data: "):
+                                data_str = line[6:]  # Remove "data: " prefix
+                                if data_str.strip() == "[DONE]":
+                                    break
                                 try:
-                                    data = json.loads(line)
-                                    token = data.get("response", "")
+                                    data = json.loads(data_str)
+                                    delta = data.get("choices", [{}])[0].get("delta", {})
+                                    token = delta.get("content", "")
                                     if token:
                                         yield token
-                                    # Check if done
-                                    if data.get("done", False):
-                                        break
                                 except json.JSONDecodeError:
                                     continue
-                            # Yield empty string to keep connection alive if needed
-                            # This helps with buffering issues
-            except httpx.ConnectError:
-                raise RuntimeError(
-                    f"Cannot connect to Ollama at {self.url}. "
-                    f"Is Ollama running? Start it with: ollama serve"
-                )
-            except Exception as e:
-                raise RuntimeError(f"LLM streaming failed: {str(e)}")
-        elif self.backend == "groq":
-            if not self.api_key:
-                raise RuntimeError(
-                    "Groq API key not configured. Set GROQ_API_KEY environment variable. "
-                    "Get a free API key at https://console.groq.com"
-                )
-            if not self.model:
-                raise RuntimeError("Groq model not configured. Set GROQ_MODEL environment variable.")
-            
+        except httpx.HTTPStatusError as e:
+            error_detail = "Unknown error"
             try:
-                async with httpx.AsyncClient(timeout=300.0) as client:
-                    async with client.stream(
-                        "POST",
-                        "https://api.groq.com/openai/v1/chat/completions",
-                        headers={
-                            "Authorization": f"Bearer {self.api_key}",
-                            "Content-Type": "application/json"
-                        },
-                        json={
-                            "model": self.model,
-                            "messages": [
-                                {"role": "user", "content": prompt}
-                            ],
-                            "temperature": temperature,
-                            "stream": True
-                        }
-                    ) as response:
-                        response.raise_for_status()
-                        async for line in response.aiter_lines():
-                            if line:
-                                # Groq uses Server-Sent Events format
-                                if line.startswith("data: "):
-                                    data_str = line[6:]  # Remove "data: " prefix
-                                    if data_str.strip() == "[DONE]":
-                                        break
-                                    try:
-                                        data = json.loads(data_str)
-                                        delta = data.get("choices", [{}])[0].get("delta", {})
-                                        token = delta.get("content", "")
-                                        if token:
-                                            yield token
-                                    except json.JSONDecodeError:
-                                        continue
-            except httpx.HTTPStatusError as e:
-                error_detail = "Unknown error"
-                try:
-                    error_json = e.response.json()
-                    error_detail = error_json.get("error", {}).get("message", str(error_json))
-                except:
-                    error_detail = e.response.text
-                raise RuntimeError(f"Groq API streaming error: HTTP {e.response.status_code} - {error_detail}")
-            except Exception as e:
-                raise RuntimeError(f"Groq API streaming failed: {str(e)}")
-        else:
-            raise RuntimeError(f"Streaming not supported for backend: {self.backend}")
+                error_json = e.response.json()
+                error_detail = error_json.get("error", {}).get("message", str(error_json))
+            except:
+                error_detail = e.response.text
+            raise RuntimeError(f"Groq API streaming error: HTTP {e.response.status_code} - {error_detail}")
+        except Exception as e:
+            raise RuntimeError(f"Groq API streaming failed: {str(e)}")

backend/api/services/metadata_extractor.py

@@ -24,10 +24,8 @@ class MetadataExtractor:
     
     def __init__(self, llm_client: Optional[LLMClient] = None):
         self.llm = llm_client or LLMClient(
-            backend=os.getenv("LLM_BACKEND", "ollama"),
-            url=os.getenv("OLLAMA_URL"),
             api_key=os.getenv("GROQ_API_KEY"),
-            model=os.getenv("OLLAMA_MODEL", "llama3.1:latest")
+            model=os.getenv("GROQ_MODEL")
         )
     
     async def extract_metadata(

backend/api/services/rule_enhancer.py

@@ -16,10 +16,8 @@ class RuleEnhancer:
     
     def __init__(self, llm_client: Optional[LLMClient] = None):
         self.llm = llm_client or LLMClient(
-            backend=os.getenv("LLM_BACKEND", "ollama"),
-            url=os.getenv("OLLAMA_URL"),
             api_key=os.getenv("GROQ_API_KEY"),
-            model=os.getenv("OLLAMA_MODEL", "llama3.1:latest")
+            model=os.getenv("GROQ_MODEL")
         )
     
     async def enhance_rule(

backend/api/storage/analytics_store.py

@@ -21,9 +21,13 @@ try:
     from supabase import Client, create_client
 
     SUPABASE_AVAILABLE = True
-except ImportError:
+except (ImportError, Exception) as e:
+    # Handle both ImportError and other exceptions (e.g., websockets.asyncio issues)
     Client = None  # type: ignore
     SUPABASE_AVAILABLE = False
+    # Only log at debug level to avoid noise - this is expected in some deployments
+    import logging
+    logging.getLogger(__name__).debug(f"Supabase import failed: {e}")
 
 logger = logging.getLogger(__name__)
 
@@ -63,15 +67,12 @@ class AnalyticsStore:
         
         if not SUPABASE_AVAILABLE:
             raise RuntimeError(
-                "Supabase package not installed. Install with: pip install supabase\n"
-                "AnalyticsStore requires Supabase - SQLite fallback has been removed."
+                "Supabase package not installed. Install with: pip install supabase"
             )
         
         if not supabase_url or not supabase_key:
             raise RuntimeError(
-                "Supabase credentials are required!\n"
-                "Set SUPABASE_URL and SUPABASE_SERVICE_KEY in your .env file.\n"
-                "AnalyticsStore requires Supabase - SQLite fallback has been removed."
+                "Supabase credentials required. Set SUPABASE_URL and SUPABASE_SERVICE_KEY."
             )
         
         self.use_supabase = True  # Always True - no fallback
@@ -110,9 +111,8 @@ class AnalyticsStore:
         except Exception as exc:
             logger.error(f"❌ Failed to initialize Supabase client for analytics: {exc}")
             raise RuntimeError(
-                f"Failed to initialize Supabase client: {exc}\n"
-                "Make sure SUPABASE_URL and SUPABASE_SERVICE_KEY are correct.\n"
-                "AnalyticsStore requires Supabase - SQLite fallback has been removed."
+                f"Failed to initialize Supabase client: {exc}. "
+                "Verify SUPABASE_URL and SUPABASE_SERVICE_KEY are correct."
             ) from exc
 
     def _quick_table_check(self):
@@ -192,8 +192,7 @@ class AnalyticsStore:
             )
             # Re-raise - no SQLite fallback
             raise RuntimeError(
-                f"Failed to insert into Supabase table '{table}': {error_msg}\n"
-                "AnalyticsStore requires Supabase - SQLite fallback has been removed."
+                f"Failed to insert into Supabase table '{table}': {error_msg}"
             ) from exc
 
     def _supabase_simple_select(

backend/mcp_server/common/logging.py

@@ -51,7 +51,13 @@ def _get_analytics_store() -> Optional["AnalyticsStore"]:
     try:
         _analytics_store = AnalyticsStore()
     except RuntimeError as exc:
-        logger.warning("Analytics disabled: %s", exc)
+        # Only log at warning level if credentials are configured (actual error)
+        # Otherwise log at debug level (expected when Supabase is not configured)
+        import os
+        if os.getenv("SUPABASE_URL") and os.getenv("SUPABASE_SERVICE_KEY"):
+            logger.warning("Analytics disabled: %s", str(exc).split('\n')[0])  # Only first line
+        else:
+            logger.debug("Analytics disabled: %s", str(exc).split('\n')[0])
         _analytics_failed = True
         _analytics_store = None
     except Exception as exc:  # pragma: no cover - unexpected failures

docker-commands.ps1

@@ -0,0 +1,56 @@
+# IntegraChat Docker Helper Commands for PowerShell
+
+# Function to start/restart the container
+function Start-IntegraChat {
+    Write-Host "Starting IntegraChat container..." -ForegroundColor Green
+    
+    # Check if container exists
+    $exists = docker ps -a --filter "name=integrachat" --format "{{.Names}}"
+    
+    if ($exists -eq "integrachat") {
+        Write-Host "Container exists. Stopping and removing..." -ForegroundColor Yellow
+        docker stop integrachat 2>$null
+        docker rm integrachat 2>$null
+    }
+    
+    # Run the container
+    docker run -d --name integrachat `
+        -p 7860:7860 `
+        -p 8000:8000 `
+        -p 8900:8900 `
+        -e DOCKER_CONTAINER=1 `
+        integrachat:latest
+    
+    if ($LASTEXITCODE -eq 0) {
+        Write-Host "`n✅ Container started!" -ForegroundColor Green
+        Write-Host "`nAccess services:" -ForegroundColor Cyan
+        Write-Host "  • Gradio UI: http://localhost:7860"
+        Write-Host "  • FastAPI:   http://localhost:8000"
+        Write-Host "  • MCP Server: http://localhost:8900"
+        Write-Host "`nView logs: docker logs -f integrachat" -ForegroundColor Yellow
+    }
+}
+
+# Function to stop the container
+function Stop-IntegraChat {
+    Write-Host "Stopping IntegraChat container..." -ForegroundColor Yellow
+    docker stop integrachat
+}
+
+# Function to view logs
+function Show-IntegraChatLogs {
+    docker logs -f integrachat
+}
+
+# Function to rebuild
+function Rebuild-IntegraChat {
+    Write-Host "Rebuilding IntegraChat..." -ForegroundColor Green
+    docker stop integrachat 2>$null
+    docker rm integrachat 2>$null
+    docker build -t integrachat:latest .
+    Start-IntegraChat
+}
+
+# Export functions
+Export-ModuleMember -Function Start-IntegraChat, Stop-IntegraChat, Show-IntegraChatLogs, Rebuild-IntegraChat
+

docker-compose.yml

@@ -0,0 +1,33 @@
+services:
+  integrachat:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    container_name: integrachat
+    ports:
+      - "7860:7860"  # Gradio UI
+      - "8000:8000"  # FastAPI
+      - "8900:8900"  # MCP Server
+    environment:
+      - API_PORT=8000
+      - MCP_PORT=8900
+      - GRADIO_PORT=7860
+      - DOCKER_CONTAINER=1
+      # Add your environment variables here or use env_file
+      # - SUPABASE_URL=${SUPABASE_URL}
+      # - SUPABASE_SERVICE_KEY=${SUPABASE_SERVICE_KEY}
+      # - GROQ_API_KEY=${GROQ_API_KEY}
+      # - OLLAMA_BASE_URL=${OLLAMA_BASE_URL}
+    env_file:
+      - .env  # Optional: load from .env file if it exists
+    volumes:
+      # Optional: mount logs directory for persistence
+      - ./logs:/app/logs
+    restart: unless-stopped
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 60s
+

docker-entrypoint.sh

@@ -19,8 +19,11 @@ log "API_PORT=${API_PORT}, MCP_PORT=${MCP_PORT}, GRADIO_PORT=${GRADIO_PORT}"
 
 cleanup() {
     log "Received termination signal. Stopping services..."
-    # Kill child processes
-    kill "${MCP_PID}" "${API_PID}" "${GRADIO_PID}" "${TAIL_PID}" 2>/dev/null || true
+    # Kill child processes (only if they exist)
+    [ -n "${MCP_PID:-}" ] && kill "${MCP_PID}" 2>/dev/null || true
+    [ -n "${API_PID:-}" ] && kill "${API_PID}" 2>/dev/null || true
+    [ -n "${GRADIO_PID:-}" ] && kill "${GRADIO_PID}" 2>/dev/null || true
+    [ -n "${TAIL_PID:-}" ] && kill "${TAIL_PID}" 2>/dev/null || true
     wait || true
     log "All services stopped. Exiting."
 }

env.example

@@ -11,28 +11,19 @@ SUPABASE_SERVICE_KEY=your_service_role_key_here
 POSTGRESQL_URL=postgresql://user:password@host:port/database
 
 # =============================================================
-# LLM CONFIGURATION
+# LLM CONFIGURATION (Groq Only)
 # =============================================================
-# Backend selection: "ollama" (local) or "groq" (cloud API)
-# For Hugging Face Spaces, use "groq"
-LLM_BACKEND=groq
-
-# Option 1: Using Groq API (recommended for Hugging Face Spaces)
 # Get free API key at https://console.groq.com
 GROQ_API_KEY=your_groq_api_key_here
-GROQ_MODEL=llama-3.1-70b-versatile
-
-# Option 2: Using local Ollama (for local development)
-# OLLAMA_URL=http://localhost:11434
-# OLLAMA_MODEL=llama3.1:latest
+GROQ_MODEL=llama-3.1-8b-instant
 
 # =============================================================
 # MCP SERVER CONFIG
 # =============================================================
-# Legacy FastAPI endpoints (remove once all callers use the unified MCP server)
-RAG_MCP_URL=http://localhost:8001
-WEB_MCP_URL=http://localhost:8002
-ADMIN_MCP_URL=http://localhost:8003
+# Unified MCP server endpoints (running on port 8900)
+RAG_MCP_URL=http://localhost:8900/rag
+WEB_MCP_URL=http://localhost:8900/web
+ADMIN_MCP_URL=http://localhost:8900/admin
 
 # Unified MCP server identifier (namespaced tools)
 MCP_SERVER_ID=integrachat

run-docker.ps1

@@ -0,0 +1,59 @@
+# PowerShell script to run IntegraChat Docker container
+
+# Stop and remove existing container if it exists
+Write-Host "Checking for existing container..." -ForegroundColor Yellow
+$existing = docker ps -a --filter "name=integrachat" --format "{{.Names}}" 2>&1
+if ($existing -eq "integrachat") {
+    Write-Host "Removing existing container (force)..." -ForegroundColor Yellow
+    # Use -f to force remove (stops and removes in one command)
+    docker rm -f integrachat 2>&1 | Out-Null
+    Start-Sleep -Seconds 1
+}
+
+# Build the image
+Write-Host "Building Docker image (this may take a few minutes)..." -ForegroundColor Green
+Write-Host "Progress will be shown below..." -ForegroundColor Gray
+docker build -t integrachat:latest .
+
+if ($LASTEXITCODE -ne 0) {
+    Write-Host "Build failed! Check the error messages above." -ForegroundColor Red
+    exit 1
+}
+Write-Host "Build completed successfully!" -ForegroundColor Green
+
+# Check if .env file exists
+if (-not (Test-Path .env)) {
+    Write-Host "Warning: .env file not found. Creating from env.example..." -ForegroundColor Yellow
+    if (Test-Path env.example) {
+        Copy-Item env.example .env
+        Write-Host "Created .env file. Please update it with your configuration." -ForegroundColor Yellow
+    } else {
+        Write-Host "Error: env.example not found. Cannot create .env file." -ForegroundColor Red
+        exit 1
+    }
+}
+
+# Run the container
+Write-Host "Starting container..." -ForegroundColor Green
+docker run -d --name integrachat `
+    -p 7860:7860 `
+    -p 8000:8000 `
+    -p 8900:8900 `
+    --env-file .env `
+    -e DOCKER_CONTAINER=1 `
+    integrachat:latest
+
+if ($LASTEXITCODE -eq 0) {
+    Write-Host "Container started successfully!" -ForegroundColor Green
+    Write-Host ""
+    Write-Host "Access services:" -ForegroundColor Cyan
+    Write-Host "  - Gradio UI: http://localhost:7860" -ForegroundColor White
+    Write-Host "  - FastAPI: http://localhost:8000" -ForegroundColor White
+    Write-Host "  - MCP Server: http://localhost:8900" -ForegroundColor White
+    Write-Host ""
+    Write-Host "View logs: docker logs -f integrachat" -ForegroundColor Yellow
+} else {
+    Write-Host "Failed to start container!" -ForegroundColor Red
+    exit 1
+}
+