Optimize

Fix your prompt from the calls it failed.

Point the optimizer at real calls, let it diagnose what went wrong, and generate a targeted rewrite. Review the exact diff, then sync it to your test agent or live in Vapi.

Optimize a promptBook a demo
app.rubrichq.io/prompt-optimizer
N
Nooruz zaman's Orga…
Owner
Platform
Dashboard
Test Agents
Co-Pilot
Metrics
Analytics
Prompt Optimizer
Evaluation
Live Simulations
Batch
Observability
Conversations
Notifications
Settings
N
Nooruz zaman
nooruzzonline@gmail.com
Prompt Optimizerv6 · diff
Agent: Billing Bot
Last updated 2m ago · diagnosed from 5 calls
Optimized

The rewrite adds an explicit empathy step before explaining disputed charges, requires confirming resolution, and reads the SSN disclosure verbatim — the three gaps behind the failing calls.

Show only changesCopy Optimized
Current InstructionsOptimized Instructions
1You are Billing Bot, a billing support agent.
1You are Billing Bot, a billing support agent.
2
2
3## Tone
3## Tone
4Be polite and professional.
4+Be warm, patient, and proactive.
5
5
6## When the caller disputes a charge
6## When the caller disputes a charge
7Explain the charge to the caller.
7+Acknowledge the frustration first, then explain the charge.
8+Confirm the issue is resolved before ending the call.
8## Identity verification
9## Identity verification
9Verify the account.
10+Read the SSN disclosure verbatim before account access.
e.g. Warmer greeting + confirm resolution
RejectSync with Test AgentSync with vendor

Diagnosed from real calls

Pick the calls that failed (or the ones that went well) and the optimizer finds the exact gaps to fix.

Targeted rewrite

The rewrite addresses each diagnosed gap and is tuned to your model — not a blind regeneration.

Reviewed, then synced

See a word-level diff and the score lift before anything ships — then push to your test agent or Vapi.

Diagnosis

It tells you what to fix before it rewrites

Select up to five calls and the optimizer reads their transcripts and metrics to surface concrete gaps — each tied to a failing metric and the calls that prove it. Edit, drop, or add gaps to steer the rewrite.

Every gap maps to a real metric and verdict
Cited evidence calls behind each finding
Add your own constraints in plain English
Nothing is rewritten until you approve the gaps
Diagnosis3 gaps

Across 5 calls the agent stayed defensive, skipped a required disclosure, and never confirmed resolution.

failDe-escalation handling

Acknowledge the caller’s frustration before reasoning.

Run #128call_9f3a
failCompliance disclosure

Read the SSN disclosure verbatim before account access.

call_8x4r
weakConfirms resolution

Ask if the issue is resolved before closing the call.

Run #131call_2k9q
Re-analyzeGenerate rewrite

Two ways to optimize

Fix what failed, or reinforce what works

Start from your worst calls to patch a regression, or from your best calls to make the agent reliably reproduce them.

failure_driven

Failure-driven

Learns from failures. Diagnoses what the instruction is missing on calls where metrics failed or were weak, then rewrites it to close those gaps — driven primarily by the failed metrics.

Patch the de-escalation regression in v6
exemplar_driven

Exemplar-driven

Learns from exemplars (good calls). Extracts the strong behaviours from both transcripts and metrics, and rewrites the instruction so the agent reliably reproduces them.

Make every call close like our best ones

More in Optimize

Every change is versioned, reversible, and one click from your live agent.

History

A version for every optimization

Each rewrite is committed as a labeled version with its diagnosis, diff, and status. See exactly which one is current on the agent — and reset to any earlier version in one click.

Labeled versionsFull diff per versionReset to any point
Optimization historyBilling Bot
v6OptimizedView
Empathy + confirm resolution
v5CurrentView
Tighter knowledge-base scope
v4RejectedViewReset
Conversational tone
v3MergedViewReset
Initial optimization
Apply the rewrite
Sync with Test Agent
Make it current in RubricHQ
Sync with vendor
Push live to Vapi or Retell
Reject
Discard and keep the current prompt
Apply & sync

One click from review to live

Approve a rewrite and it becomes current on your RubricHQ test agent. When you're ready, push the same prompt straight to your live agent in Vapi or Retell — or reject it and nothing changes.

Test agent firstPush to Vapi / RetellReject anytime

How it works

From failing calls to a shipped fix

1

Diagnose

Pick up to 5 calls. The optimizer reads them and surfaces the gaps to fix, each tied to a metric.

2

Generate the rewrite

A targeted rewrite is generated to close each diagnosed gap, tuned to your model and constraints.

3

Review & apply

Read the word-level diff and score lift, then sync to your test agent — or push live to Vapi or Retell.

Stop guessing at your prompt.

Diagnose from real calls, generate a targeted fix, and ship it with confidence.

Optimize a promptBook a demo