IntelliRoute
User Workspace
User Workspace
Product · IntelliRoute

Welcome to IntelliRoute

A distributed LLM control plane that routes every request across providers with intent-aware policies, global quotas, and full-stack observability—so teams ship faster without babysitting models.

Multi-Model Cost-Aware Observable
Routing
Intent + policy engine
Resilience
Fallback & brownout-aware
Governance
Quotas · cost · feedback
Demo mode — no sign-in required
✨

User Workspace

Prompt studio and responses tuned for clarity. IntelliRoute picks the model path—you focus on the task.

◎

Admin Console

Live traces, provider health, quotas, costs, and ops events for engineers running the control plane.

IntelliRoute Studio

What would you like to solve?

Describe your task in natural language. We route across providers and tune for speed, quality, and cost—no manual model picking.

Suggestions
Working on it
Understanding your request…
Understand Route Generate
Your prompt
Handled by IntelliRoute

Was this helpful?

Force-fail demo toggles Gemini; reasoning/code then fall back to Groq and mocks.

Control plane overview

IntelliRoute — health, traffic, cost, and providers at a glance

Gateway Checking…
Providers (circuits) —
Quota leader —
Brownout —
Discovery —
Budget scope No team/workflow scope in this UI
Total requests
—
From cost summary (tenant)
Success rate
—
Mean EMA across providers
Avg latency
—
Mean EMA across providers
Total spend
—
Premium cost share
—
Est. from provider spend mix
Active providers
—
Registry routable

Queue traffic

Sampled every 2s while this tab is open · depth = high + medium + low

Queue: —

Cost snapshot

Spend by provider (same source as Live ops)

Provider snapshot

—

Provider Circuit Routable Latency Success Quality Anomaly

Recent events

Same stream as Live ops · newest first

Cost & quotas

Tenant spend, provider mix, budget headroom, premium exposure, and quota pressure — gateway summary plus read-only cost-tracker governance APIs.

Spend summary

Snapshot from gateway /v1/cost/summary and cost tracker rollups where available.

Budgets & usage

Tenant cap, team/workflow budgets, and utilization from the cost tracker.

Routing impact

Signals that correlate with degrade, fallback, shedding, brownout, and policy budget hints.

Limits & governance events

Cost-tracker alerts, workflow pressure hints, premium caps, plus recent ops markers.

Selection detail

Click a row in Spend summary to drill into tenant, provider, team, or workflow scope.

Tenant scope selected by default.

Live trace

This view shows the latest captured routing trace from user or admin sends when available, including completion payload, optional router /decide snapshot, and policy fields.

No trace available yet. Routing details will appear here once trace data is available.

Routing journey

How this request moved through IntelliRoute — from intake to model output.

Policy evaluation

Brownout & routing

Route candidates

Completion

Model output

Providers

Provider intelligence view for health, routability, quality, anomalies, and spend signals.

Provider leaderboard

Provider Status Latency Success Quality Anomaly Cost Signal

Provider detail

Select a provider row to inspect details.

Live ops

Operational status, system events, and control-plane signals — refreshed every 2s from existing gateway/router/health endpoints.

Poll 2s
Events —

System event stream

Newest first. Timeline uses existing UI events plus derived polling transitions.

Brownout · queue · load

Current overload posture and queue pressure signals.

Discovery · replicas · health

Leader election, discovery state, stale providers, and circuits.

Selected event

Context for the event selected in the stream.

Select an event to inspect metadata.

User feedback insights

Thumbs-up/down feedback with optional comments. AI analysis runs only when you click the button and uses the same IntelliRoute provider path as completions.

Recent feedback (newest first)

Search and filters apply to rows loaded from the server (up to 500).

No feedback yet.

AI analysis

Cached result
No analysis yet. Choose a sample size and click Analyze Feedback with AI.

                                    

                                    

                                

                            

                        

                    

                


                

                    

                        

                            

                                
Experiments

                                
Replay-eval benchmark console for IntelliRoute vs baseline routing policies across scenarios.

                            

                            

                                
                                
                            

                        


                        

                            
                            
                            
Source: waiting for experiment artifact…

                        


                        

                            
No experiment results loaded

                            
Load a replay artifact summary to compare IntelliRoute against baseline policies.

                            

                                
                                
                            

                            
Expected artifact: aggregate_summary.json from artifacts/matrix_runs/<timestamp>/.

                        


                        


                        

                            

                                
Policy comparison

                                
Success, latency, cost, premium usage, and reliability trade-offs for selected scenario.

                            

                            

                                

                                    
                                        

                                            Policy
                                            Success
                                            Avg latency
                                            P95 latency
                                            Total cost
                                            Premium usage
                                            Fallback
                                            Reroute/downgrade
                                            Reject
                                        

                                    
                                    
                                

                            

                        


                        

                            

                                

                                    
Comparison visuals

                                    
Compact bars for quick metric scanning by policy.

                                

                                

                            

                            

                                

                                    
Findings

                                    
Derived only from currently loaded experiment rows.

                                

                                

                            

                        


                        

                            

                                
Run detail

                                
Artifact metadata and scenario context.

                            

                            

                        

                    

                

            

        


        

            

            

                

                    

                        
Prompt history

                        
Stored in this browser tab only (session).