Voice AI

Speak to your app's test-derived memory.

The hands-free interface to TestRelic's application memory layer. Ask “which checkout selectors broke this week?” and get a spoken answer plus a generated dashboard — grounded in real test runs.

hands-freefrom any standup
spokenanswer + visual response
0typing required
Voice AI · standup · 9:02 AM

Hey TestRelic — what broke overnight?

Spoken answer · 3 failures categorized · last 12h

3

Failures

100%

Categorized

3

Linked traces

checkout.spec.ts
P0
auth.spec.ts
P1
search.spec.ts
P2

Voice answer reads the same memory layer Cursor and Claude Code use over MCP. One source of truth.

How Voice AI Works

Speak naturally — TestRelic AI understands the context of your Playwright data

Step 1

Speak Your Question

Press the mic button or use the hotkey, then ask about failures, trends, or anything in your test data

Step 2

AI Processes Context

TestRelic AI translates your voice into a structured query against your Playwright analytics in real time

Step 3

Get Instant Answer

Receive a spoken summary plus a visual response — a chart, table, dashboard, or trace link

Example Voice Queries

Real queries your QA team can ask during standups, triage, or while coding

Which Playwright tests failed in the last CI run?

3 tests failed in the checkout suite — all related to a Stripe API timeout. Traces are linked.

Show me flaky tests from this week

Found 7 flaky tests. Top offender: login_oauth — failed 4 of 12 runs. Retry analysis suggests a timing issue.

Compare test duration between main and feature-xyz

feature-xyz runs 18 % slower overall. The product-search spec accounts for most of the increase (4.2 s ? 6.8 s).

Create a dashboard for this sprint's test health

Dashboard created with pass rate, flaky trend, failure categories, and top failing specs. Shared to #qa-team.

Fits Your Existing Workflow

Dashboard Generation

Say 'Create a dashboard for this sprint' and Voice AI builds it from your Playwright data automatically.

Failure Triage

During stand-ups, ask 'What broke overnight?' and get a spoken summary with linked traces and screenshots.

Continuous Feedback

Voice AI learns from corrections. Say 'That test is not flaky, it's an env issue' and it updates the categorization.

Built for every role on the team

Voice AI speaks the language of QA, engineering, release, and product alike.

QA Engineer

Prompt

Hey TestRelic — what's flaky this week?

Spoken summary of top 5 flaky suites with retry counts, plus a generated dashboard sent to Slack.

Engineering Lead

Prompt

Compare main vs feature-xyz test duration

Voice answer: 'feature-xyz runs 18% slower' — followed by a chart broken down by spec.

Release Manager

Prompt

Build a release readiness summary for the 4pm meeting

Polished spoken brief + slide deck with pass rates, regressions, and known flakes.

Product Manager

Prompt

Walk me through what broke this sprint

Conversational summary tied to the user-impact map — no diving into logs or CI.

hands-free
from any device

Talk during standups, while coding, on the move

Multimodal
responses

Spoken answer plus a chart, dashboard, or trace link — you hear it and see it

MCP
shared memory

Voice answers come from the same memory layer your AI agents read

Your voice queries stay private

Voice input is transcribed and processed against your tenant's test memory only. Audio is not retained after transcription; transcripts and outputs are never used to train shared models.

Talk to your tests. Listen to your app.

Voice AI is available on the Growth plan. Start with a 14-day free trial — no credit card required.