Learning Dashboard
Knowledge base health, learning pipeline status, and recent ingestions
🎯 Learning Pipeline
Knowledge base health, learning pipeline status, and recent ingestions. Content flows: Upload → Chunk → Embed → Index → Available.
Data Collection
Upload and organize your business-specific training data
CompletePreprocessing
Clean and prepare data for optimal training results
CompleteModel Training
Train the AI model with your customized dataset
In ProgressEvaluation
Test model performance and accuracy metrics
PendingDeployment
Deploy your custom model to production
Pending📈 Learning & Knowledge Metrics
—
Knowledge Items
—
Recent (7d)
—
Queue Depth
—
Pipeline Health
🎯 Insights
Live insights from your learning data and usage
Loading insights…
📋 Training Logs
Training Session Management
Monitor learning sessions (legacy); use Learning Activity for recent events
🔍 Filters
📋 Active Sessions
Loading...
Loading sessions...
Knowledge Base Management
Manage and curate Bob's knowledge base with full CRUD operations
🔍 Search & Filters
0
📚 Knowledge Items
Loading...
Loading knowledge items...
⭐ Quality Management
Monitor and manage knowledge quality scores, approval workflows, and quality overrides.
Average Quality Score
--
Pending Review
--
Approved Items
--
Rejected Items
--
Tenant Overview
Monitor and manage learning across all businesses - see who's learning what and track progress
📊 System Overview
--
Total Businesses
--
Active Training
--
Total Knowledge Items
--
Training Sessions
--
Learning Goals
--
Memory Items
🏢 Business Learning Status
View detailed learning metrics and training status for each business
Loading businesses...
📈 Cross-Business Comparison
Compare learning progress and knowledge accumulation across businesses
Loading comparison chart...
Content Sources
Documents, crawled sites, learned-from-conversations by source
📤 Upload Training Data
Upload your business-specific data to train Bob AI. Supported formats include conversation logs, FAQ documents, product information, and customer service transcripts.
Drag & Drop Files Here
or click to browse files
JSON
CSV
TXT
PDF
XLSX
📚 Content Sources
Knowledge items grouped by source type. Use Knowledge Base for full CRUD.
Loading content sources...
Content Processing Pipeline
Upload → Chunk → Embed → Index → Available
📊 Pipeline Status
Loading pipeline status...
⚙️ Preprocessing Pipeline
Content processing pipeline configuration is managed per-tenant via the API. Pipeline steps: Upload → Chunk → Embed → Index → Available.
Pipeline configuration loads from tenant settings. Use the Knowledge Base tab to manage content.
📊 Data Quality Metrics
—
Total Examples
—
Data Quality Score
247
Duplicates Found
2,847
Avg. Tokens/Example
🔍 Data Preview
Sample of processed data showing transformations applied
Example 1: Customer Support Query
ORIGINAL
PROCESSED
Learning Activity
Learning goals and pipeline activity — what the learning worker processed
📋 Recent Learning Activity
Pipeline events (ingestion, processing) from the learning queue. Goals are in Coaching.
Loading learning activity...
🎛️ Training Configuration
Planned. Bob uses RAG (retrieval-augmented generation), not model fine-tuning. Configure LLM providers in AI Provider Configuration.
🖥️ Resource Usage
78%
GPU Utilization
45°C
GPU Temperature
12.4 GB
Memory Usage
$24.50
Current Run Cost
📋 Training Logs
INFO
SUCCESS
INFO
SUCCESS
Response Quality Metrics
Retrieval hit rate, response satisfaction, escalation rate, knowledge gap frequency
📊 Overall Performance
96.8%
Overall Accuracy
95.2%
Precision
94.7%
Recall
94.9%
F1 Score
🎯 Per-Class Accuracy
Performance breakdown by classification category
Product Inquiry
98.2%
Support Request
96.5%
Order Status
97.1%
Returns
95.8%
Billing
94.3%
Shipping
96.9%
📈 Performance Charts
🔍 Confusion Matrix
Visual representation of prediction accuracy across classes. Darker cells indicate higher prediction counts.
Product
Support
Order
Returns
Billing
Shipping
Product
982
8
3
2
1
4
Support
12
965
15
3
2
3
Order
5
10
971
8
2
4
Returns
3
4
12
958
18
5
Billing
2
3
4
24
943
24
Shipping
6
5
10
4
6
969
AI Provider Configuration
Configured LLM providers, status, fallback chain, cost
🤖 Active LLM Provider
Primary model from tenant configuration. Change in Mission Control → Publish Room or tenant config.
Loading provider config...
Publish Configuration
Active config and deploy from Mission Control → Publish Room
📌 Active Configuration
Loading...
🌍 Environment Selection
Choose the target environment for your model deployment
Development
Testing and development environment with debug features
Low Traffic
Debug Mode
Staging
Pre-production environment for final testing
Medium Traffic
Production-like
Production
Live production environment serving real users
High Traffic
Auto-scaling
⚙️ Deployment Configuration
Method for deploying new version
Automatic capacity scaling
Endpoint for health monitoring
How often to check health
Automatic rollback trigger
New version: 100% | Old version: 0%
📊 Deployment Status
v2.1.0
Current Version
Active
99.98%
Uptime (30 days)
1,247
Requests/min
145ms
Avg Response Time
🔄 Version History
Recent deployments and version management
v2.1.0 - Current Production
Deployed
2 days ago
Success Rate
99.2%
Accuracy
96.8%
Total Requests
3.2M
v2.0.5 - Previous Version
Deployed
1 week ago
Success Rate
98.9%
Accuracy
95.2%
Total Requests
8.7M
Learning Pipeline Monitor
Real-time performance tracking and system health monitoring
⚡ Learning Pipeline Metrics
—
Queue Depth
—
Last Learning Run
—
Pipeline Health
—
Knowledge Items
📊 Response Time Distribution
98ms
P50 (Median)
187ms
P90
342ms
P95
487ms
P99
🖥️ System Health
4/4
Healthy Instances
All Healthy
45%
CPU Usage
62%
Memory Usage
99.98%
Uptime (7 days)
🔔 Active Alerts
System alerts and notifications requiring attention
⚠️ High P99 Latency
WARNING
ℹ️ High Traffic Volume
INFO
⚠️ Memory Usage Increasing
WARNING
📈 Performance Trends
Learning Analytics
Memory maps, knowledge flow, cost projections from Learning Observatory
📊 Observatory Status
Loading...
📊 Performance Over Time
💰 Cost Analysis
$1,247
Total Training Costs
$3,892
Inference Costs (30d)
$0.003
Cost per Request
-18%
Cost Reduction (vs v2.0)
📈 Usage Patterns
45.2M
Total Requests (All Time)
1.8M
Requests (Last 7 Days)
874
Peak Requests/min
2.3 hours
Peak Usage Time
2-4 PM UTC
💡 ROI Summary
Estimated savings derived from your usage data. Detailed ROI projections are available in the Analytics dashboard.
—
Est. Monthly Savings
—
ROI
—
Payback Period
🎯 Key Insights
Insights from your learning data and usage (from API)
Loading insights…
Test Your Knowledge
Test if Bob can answer questions from your knowledge base
🧪 Interactive Testing Console
Test if Bob can answer from your knowledge base. Send a message and see the response.
Testing Console
Response will appear here after testing...
📊 Performance Comparison
🧠 Cognitive Health Dashboard
Unified view of system intelligence, learning velocity, and cognitive metrics
--
Overall Cognitive Health Score
📈 Learning Velocity
Loading...
🧠 Memory Health
Loading...
📚 Knowledge Coverage
Loading...
🎯 Goal Performance
Loading...
💰 Intervention ROI
Loading...
📊 Trends Over Time
📝 Cognitive Audit Log
Loading audit log...
🌐 Cross-Tenant Pattern Recognition
Anonymized insights from patterns across all businesses
🔒
Privacy-Preserving Analysis
All data is anonymized and aggregated. No tenant-specific information is exposed.
Loading cross-tenant patterns...
⚖️ Cognitive Load Management
Monitor and manage system cognitive load to prevent overload
--
Loading...
System Cognitive Load (0-100)
0
Memories
0
Knowledge Items
0
Active Goals
0
Pending Interventions
💡 Recommendations
Loading recommendations...
🔧 Simplified Mode
Reduce cognitive load by simplifying system operations
📈 Load History
Learning Goals & Coaching
Guide Bob's learning with manual knowledge injection and learning goals
🏢 Tenant:
Loading...
📊 Coaching Analytics
Overview of learning goals and intervention effectiveness
🎯 Learning Goals
Set and track learning objectives for Bob's development
Loading learning goals...
📝 Intervention History
Review past coaching interventions and knowledge injections
Loading intervention history...
💉 Manual Knowledge Injection
Inject knowledge directly into Bob's knowledge base
💡 Intelligent Intervention Suggestions
AI-powered recommendations to improve Bob's learning
Loading intervention suggestions...
🔮 Predictive Learning Goals
AI-generated learning goals based on your business category and patterns
No predictive goals generated yet. Click "Generate Goals" to get started.
📊 Intervention Impact Tracking
Track the effectiveness of coaching interventions over time
0%
Avg Confidence Increase
0
Total Interventions
0
Successful
0
Avg Impact Score
Loading impact tracking data...
Memory Explorer
Monitor and manage Bob's multi-level memory system
📊 Memory Level Statistics
Overview of memory distribution across different levels
Level 1: Episodic Memory
0Short-term, conversation-specific memories
Avg Importance: 0.00
Level 2: Working Memory
0Active, frequently accessed information
Avg Importance: 0.00
Level 3: Semantic Memory
0Conceptual knowledge and facts
Avg Importance: 0.00
Level 4: Long-term Memory
0Stable, well-established knowledge
Avg Importance: 0.00
Level 5: Permanent Memory
0Core knowledge that persists across sessions
Avg Importance: 0.00
📋 Promotion Queue
Memories pending promotion (from GET /api/learning/memory/promotion-queue)
Loading…
🔍 Search Memory
Search by keyword (GET /api/learning/memory/search)
Enter a term and click Search.
🔄 Memory Operations
⚙️ Memory Consolidation
Trigger manual memory consolidation to optimize storage
Idle
🔗 Memory Associations Network
Visualize connections between related memories
Select a memory to view its associations
⏳ Memory Decay Management
Monitor and manage memories scheduled for decay
0
Decay Candidates
0
Decayed (30d)
0.00
Avg Importance
Loading decay candidates...
🖼️ Multi-Modal Memory
Upload and manage files (documents, images, audio, video) as memories
📁
Drag & Drop Files Here
or click to browse
Supports: PDF, DOCX, Images, Audio, Video
Loading multi-modal memories...
⏰ Temporal Memory Sequences
View temporal patterns and causal relationships between memories
Loading temporal sequences...
Loading temporal sequences...