fix: Revamp UI and debloating

EfficientAI-Docs/docs/getting-started/installation.md

@@ -8,29 +8,32 @@ sidebar_position: 1
 
 There are two ways to run the application:
 
-## Method 1: Using Docker Compose
+## Method 1: Using Docker Compose (Recommended)
 
-Start all services:
+Start all services with a single command:
 
 ```bash
 docker compose up -d
 ```
 
 This will automatically:
 
-- Build Docker images if they don't exist
-- Build the frontend during the Docker build process
+- **Pull pre-built images** from GitHub Container Registry (no build required!)
 - Start all services (database, Redis, API, worker)
 - Run database migrations automatically on startup
 
-**Note**: If you make changes to the frontend or backend code, you may need to rebuild:
+The first run will download ~4GB of images, which typically takes 1-2 minutes depending on your internet speed.
+
+### Using a Specific Version
+
+You can pin to a specific release version for stability:
 
 ```bash
-# Rebuild and restart (forces rebuild even if image exists)
-docker compose up -d --build
+# Use a specific version
+EFFICIENTAI_VERSION=1.0.0 docker compose up -d
 
-# Or rebuild without cache for a clean build
-docker compose build --no-cache api
+# Or add to your .env file for persistence
+echo "EFFICIENTAI_VERSION=1.0.0" >> .env
 docker compose up -d
 ```
 
@@ -57,10 +60,17 @@ docker compose exec api python scripts/create_api_key.py "My API Key"
 - Frontend: http://localhost:8000/
 - API Docs: http://localhost:8000/docs
 
-**Note**: The frontend is automatically built into the Docker image during the first `docker compose up -d` command. If you make frontend changes later, rebuild with:
+### Building Locally (for development)
+
+If you want to build images locally instead of pulling pre-built ones (e.g., for development):
 
 ```bash
+# Edit docker-compose.yml to uncomment the 'build' sections, then:
 docker compose up -d --build
+
+# Or rebuild without cache for a clean build
+docker compose build --no-cache api worker
+docker compose up -d
 ```
 
 ## Method 2: Using Command Line (CLI)
@@ -210,9 +220,34 @@ This will:
 
 **For Docker Compose**:
 - Docker and Docker Compose installed
+- ~4GB disk space for pre-built images
 
 **For CLI**:
 - Python 3.11+
 - Node.js 18+ and npm
 - PostgreSQL running (locally or remote)
 - Redis running (locally or remote)
+
+## Docker Images
+
+EfficientAI provides pre-built Docker images hosted on GitHub Container Registry:
+
+| Image | Description | Size |
+|-------|-------------|------|
+| `ghcr.io/efficientai-tech/efficientai-api` | API server + frontend | ~1.5GB |
+| `ghcr.io/efficientai-tech/efficientai-worker` | Celery worker with ML models | ~4GB |
+
+### Available Tags
+
+- `latest` - Most recent build from main branch
+- `x.y.z` - Specific version (e.g., `1.0.0`)
+- `x.y` - Latest patch of a minor version (e.g., `1.0`)
+
+### Manual Pull (Optional)
+
+Images are pulled automatically by `docker compose up`, but you can pre-pull them:
+
+```bash
+docker pull ghcr.io/efficientai-tech/efficientai-api:latest
+docker pull ghcr.io/efficientai-tech/efficientai-worker:latest
+```

README.md

@@ -45,26 +45,25 @@ Test quality, measure performance, & ship with confidence.
 
 There are two ways to run the application:
 
-### Method 1: Using Docker Compose
+### Method 1: Using Docker Compose (Recommended)
 
 1. **Start all services**
    ```bash
    docker compose up -d
    ```
 
    This will automatically:
-   - Build Docker images if they don't exist
-   - Build the frontend during the Docker build process
+   - Pull pre-built images from GitHub Container Registry (no build required!)
    - Start all services (database, Redis, API, worker)
    - Run database migrations automatically on startup
 
-   **Note:** If you make changes to the frontend or backend code, you may need to rebuild:
+   **Using a specific version:**
    ```bash
-   # Rebuild and restart (forces rebuild even if image exists)
-   docker compose up -d --build
+   # Pin to a specific release version
+   EFFICIENTAI_VERSION=1.0.0 docker compose up -d
 
-   # Or rebuild without cache for a clean build
-   docker compose build --no-cache api worker
+   # Or add to your .env file
+   echo "EFFICIENTAI_VERSION=1.0.0" >> .env
    docker compose up -d
    ```
 
@@ -81,6 +80,19 @@ There are two ways to run the application:
    - Frontend: http://localhost:8000/
    - API Docs: http://localhost:8000/docs
 
+#### Building Locally (for development)
+
+If you want to build images locally instead of pulling pre-built ones:
+
+```bash
+# Edit docker-compose.yml to uncomment the 'build' sections, then:
+docker compose up -d --build
+
+# Or rebuild without cache for a clean build
+docker compose build --no-cache api worker
+docker compose up -d
+```
+
 ### Method 2: Using Command Line (CLI)
 
 1. **Install the package**
@@ -161,6 +173,7 @@ There are two ways to run the application:
 
 **For Docker Compose:**
 - Docker and Docker Compose installed
+- ~4GB disk space for pre-built images
 
 **For CLI:**
 - Python 3.11+

app/api/v1/routes/agents.py

@@ -3,7 +3,7 @@
 Complete CRUD operations for test agents
 """
 from fastapi import APIRouter, Depends, HTTPException, status, Query
-from fastapi.responses import JSONResponse
+from fastapi.responses import JSONResponse, Response
 from sqlalchemy.orm import Session
 from typing import List, Optional
 from uuid import UUID
@@ -147,6 +147,48 @@ def generate_unique_agent_id(db: Session) -> str:
     )
 
 
+def get_agent_dependencies(db: Session, organization_id: UUID, agent_uuid: UUID) -> dict:
+    """Return dependency counts that block non-force delete."""
+    evaluators_count = db.query(Evaluator).filter(
+        Evaluator.agent_id == agent_uuid,
+        Evaluator.organization_id == organization_id,
+    ).count()
+
+    evaluator_results_count = db.query(EvaluatorResult).filter(
+        EvaluatorResult.agent_id == agent_uuid,
+        EvaluatorResult.organization_id == organization_id,
+    ).count()
+
+    call_recordings_count = db.query(CallRecording).filter(
+        CallRecording.agent_id == agent_uuid,
+        CallRecording.organization_id == organization_id,
+    ).count()
+
+    conversation_evaluations_count = db.query(ConversationEvaluation).filter(
+        ConversationEvaluation.agent_id == agent_uuid,
+        ConversationEvaluation.organization_id == organization_id,
+    ).count()
+
+    test_conversations_count = db.query(TestAgentConversation).filter(
+        TestAgentConversation.agent_id == agent_uuid,
+        TestAgentConversation.organization_id == organization_id,
+    ).count()
+
+    dependencies = {}
+    if evaluators_count > 0:
+        dependencies["evaluators"] = evaluators_count
+    if evaluator_results_count > 0:
+        dependencies["evaluator_results"] = evaluator_results_count
+    if call_recordings_count > 0:
+        dependencies["call_recordings"] = call_recordings_count
+    if conversation_evaluations_count > 0:
+        dependencies["conversation_evaluations"] = conversation_evaluations_count
+    if test_conversations_count > 0:
+        dependencies["test_conversations"] = test_conversations_count
+
+    return dependencies
+
+
 @router.post("", response_model=AgentResponse, status_code=status.HTTP_201_CREATED)
 async def create_agent(
     agent: AgentCreate,
@@ -380,56 +422,20 @@ async def delete_agent(
         raise HTTPException(status_code=404, detail=f"Agent {agent_id} not found")
 
     agent_uuid = db_agent.id
-
-    evaluators_count = db.query(Evaluator).filter(
-        Evaluator.agent_id == agent_uuid,
-        Evaluator.organization_id == organization_id,
-    ).count()
-
-    evaluator_results_count = db.query(EvaluatorResult).filter(
-        EvaluatorResult.agent_id == agent_uuid,
-        EvaluatorResult.organization_id == organization_id,
-    ).count()
-
-    call_recordings_count = db.query(CallRecording).filter(
-        CallRecording.agent_id == agent_uuid,
-        CallRecording.organization_id == organization_id,
-    ).count()
-
-    conversation_evaluations_count = db.query(ConversationEvaluation).filter(
-        ConversationEvaluation.agent_id == agent_uuid,
-        ConversationEvaluation.organization_id == organization_id,
-    ).count()
-
-    test_conversations_count = db.query(TestAgentConversation).filter(
-        TestAgentConversation.agent_id == agent_uuid,
-        TestAgentConversation.organization_id == organization_id,
-    ).count()
-
-    dependencies = {}
-    if evaluators_count > 0:
-        dependencies["evaluators"] = evaluators_count
-    if evaluator_results_count > 0:
-        dependencies["evaluator_results"] = evaluator_results_count
-    if call_recordings_count > 0:
-        dependencies["call_recordings"] = call_recordings_count
-    if conversation_evaluations_count > 0:
-        dependencies["conversation_evaluations"] = conversation_evaluations_count
-    if test_conversations_count > 0:
-        dependencies["test_conversations"] = test_conversations_count
+    dependencies = get_agent_dependencies(db, organization_id, agent_uuid)
 
     if dependencies and not force:
         parts = []
-        if evaluators_count > 0:
-            parts.append(f"{evaluators_count} evaluator(s)")
-        if evaluator_results_count > 0:
-            parts.append(f"{evaluator_results_count} evaluator result(s)")
-        if call_recordings_count > 0:
-            parts.append(f"{call_recordings_count} call recording(s)")
-        if conversation_evaluations_count > 0:
-            parts.append(f"{conversation_evaluations_count} conversation evaluation(s)")
-        if test_conversations_count > 0:
-            parts.append(f"{test_conversations_count} test conversation(s)")
+        if dependencies.get("evaluators"):
+            parts.append(f"{dependencies['evaluators']} evaluator(s)")
+        if dependencies.get("evaluator_results"):
+            parts.append(f"{dependencies['evaluator_results']} evaluator result(s)")
+        if dependencies.get("call_recordings"):
+            parts.append(f"{dependencies['call_recordings']} call recording(s)")
+        if dependencies.get("conversation_evaluations"):
+            parts.append(f"{dependencies['conversation_evaluations']} conversation evaluation(s)")
+        if dependencies.get("test_conversations"):
+            parts.append(f"{dependencies['test_conversations']} test conversation(s)")
 
         raise HTTPException(
             status_code=status.HTTP_409_CONFLICT,
@@ -479,5 +485,40 @@ async def delete_agent(
             },
         )
 
-    return JSONResponse(status_code=204, content=None)
+    return Response(status_code=204)
+
+
+@router.get("/{agent_id}/delete-impact")
+async def get_agent_delete_impact(
+    agent_id: str,
+    organization_id: UUID = Depends(get_organization_id),
+    db: Session = Depends(get_db)
+):
+    """Preview dependent records that would be affected by force delete."""
+    try:
+        agent_uuid = UUID(agent_id)
+        db_agent = db.query(Agent).filter(
+            and_(
+                Agent.id == agent_uuid,
+                Agent.organization_id == organization_id
+            )
+        ).first()
+    except ValueError:
+        db_agent = db.query(Agent).filter(
+            and_(
+                Agent.agent_id == agent_id,
+                Agent.organization_id == organization_id
+            )
+        ).first()
+
+    if not db_agent:
+        raise HTTPException(status_code=404, detail=f"Agent {agent_id} not found")
+
+    dependencies = get_agent_dependencies(db, organization_id, db_agent.id)
+    return {
+        "agent_id": str(db_agent.id),
+        "agent_name": db_agent.name,
+        "dependencies": dependencies,
+        "can_delete_without_force": len(dependencies) == 0,
+    }
 

app/api/v1/routes/conversation_evaluations.py

@@ -85,8 +85,6 @@ async def create_conversation_evaluation(
 2. A brief reason for your answer
 3. Additional metrics about the conversation quality:
    - Professionalism (0-1 scale)
-   - Clarity (0-1 scale)
-   - Empathy (0-1 scale, if applicable)
    - Problem Resolution (0-1 scale, if applicable)
    - Overall Quality (0-1 scale)
 4. An overall score (0.0 to 1.0) representing how well the agent performed
@@ -97,8 +95,6 @@ async def create_conversation_evaluation(
     "objective_achieved_reason": "brief explanation",
     "additional_metrics": {{
         "professionalism": 0.0-1.0,
-        "clarity": 0.0-1.0,
-        "empathy": 0.0-1.0,
         "problem_resolution": 0.0-1.0,
         "overall_quality": 0.0-1.0
     }},
@@ -151,6 +147,9 @@ async def create_conversation_evaluation(
         objective_achieved = bool(evaluation_data.get("objective_achieved", False))
         objective_achieved_reason = evaluation_data.get("objective_achieved_reason", "")
         additional_metrics = evaluation_data.get("additional_metrics", {})
+        # Explicitly remove deprecated dimensions so they are not evaluated/stored.
+        additional_metrics.pop("clarity", None)
+        additional_metrics.pop("empathy", None)
         overall_score = float(evaluation_data.get("overall_score", 0.0))
 
         # Calculate audio metrics if enabled and audio is available