Confused on day one

Nov 18

Spent 40 minutes trying to set this up and gave up. Why is this so complicated?

Used to love it

Nov 17

I've been a happy user for two years but the notifications are unbearable. Switched to a competitor last week.

Please add offline

Nov 16

Almost perfect. Just need offline mode for my commute. Would pay extra for it.

// the briefing · vol. 1 · issue 47filed 2m ago

agent dispatch — analyst on duty

Ask your reviews,
get the answer.

The agent reads every App Store and Google Play review you have — then files you a written briefing. Churn signals, feature requests, sentiment trends. Cited, not summarised.

Brief #00471

Why are users churning this month?

answer in 1.4s

Brief the agent on your app ▸ Watch a 60-second demo

connects to App Store + Google Play · runs on Claude, GPT-5, Gemini · free to start · no card

Put a question to it.

Pick a prompt or type your own. The agent answers in its own voice — every claim traces back to a real review.

★★★★★

awaiting first question

Every signal worth catching — before you find out the slow way.

The agent doesn't scan keywords. It clusters intent, tracks longitudinal drift, and names the review behind every claim. Six things it watches continuously, with the actual quote for each.

// sentiment drift01 / 06
How feelings shift, version by version
The agent tracks net sentiment per release and surfaces the words that moved the needle.
↳ exhibit
v3 broke the joy
Nov 14
v2 was beautiful. v3 feels cluttered and slow. Bring back the old layout.
Devon R.· 🇨🇦 CA· iOS
// churn precursor02 / 06
The early-warning phrases users say first
Past-tense affection, competitor names, "used to" — the agent flags the language people use before they leave.
↳ exhibit
Switched
Nov 12
I used to recommend this to everyone. Moved to Linear last week. Sorry.
Lena P.· 🇩🇪 DE· Play
// feature demand03 / 06
What 218 people asked for, ranked
Feature requests clustered by intent — not just keyword frequency. The agent dedupes synonyms and surfaces real demand.
↳ exhibit
Nov 16
Almost perfect. Just need offline mode for my commute. Would pay extra.
Aiko M.· 🇯🇵 JP· iOS
// competitor name-drop04 / 06
Who your users mention when they leave
Twelve named Notion this month. Eight named Linear. The agent counts attention, not just sentiment.
↳ exhibit
Trying Notion
Nov 11
Going to try Notion AI for a few weeks. Will come back if you fix sync.
Theo S.· 🇫🇷 FR· Play
// reply draft05 / 06
An empathetic response to every one-star
Tone-matched, brand-safe drafts you can edit and ship — never a copy-paste apology generator.
↳ exhibit
Embarrassing
Nov 17
App crashed three times during my presentation. I lost the room. Unbelievable.
Marcus T.· 🇬🇧 UK· Play
// locale gap06 / 06
Where reviews are getting worse, by country
50+ languages parsed natively. The agent breaks down regional rating velocity before you find out the slow way.
↳ exhibit
Nov 15
App is completely broken in Hindi. Date picker shows random characters.
Pranav K.· 🇮🇳 IN· Play

From a cold connection to a Slack ping, under a minute.

Six entries from a real morning. The agent works in the open — you can read what it did, why, and which reviews it cited.

09:14
⌁
connect
Connected Clarity Notes — App Store (US) and Google Play. Started ingestion immediately.
↳ two stores · five locales
09:14
▤
ingest
Pulled 14,832 reviews going back twelve months. Took 23 seconds. Re-indexed nightly from here.
↳ 14,832 reviews · 28s
09:31
★
find
Flagged a sync-conflict cluster. 1,247 reviews mention losing notes after switching devices; the spike begins the day v3.2 shipped.
↳ cluster #c-0042 · confidence 0.91
10:02
★
find
iOS users complain about this four times more than Android. The phrase "lost my notes" appears in 23% of one-star reviews this period.
↳ platform skew · 4.0× iOS
10:18
✎
draft
Drafted ten reply templates for the top affected reviews. Each one acknowledges the loss, references the v3.2.1 fix, and offers thirty days of credit. Ready for your review.
↳ 10 drafts · awaiting approval
10:24
◈
alert
Pinged #product-feedback in Slack with a summary, the top three quotes, and links back to source reviews. Engineering already opened a ticket.
↳ slack · #product-feedback

every claim links back to source reviews · no black boxes

// runs on the model you already trust

Claude
GPT-5
Gemini 2.5
Llama 3
Mistral
your own

no lock-in · your data stays yours

What teams have filed back — about the agent itself.

These are real reviews of the agent, written by the teams that brief it. We left them as they came in.

We caught churn in Southeast Asia

Nov 19

Before, I'd manually read reviews every Friday. Now I just ask. Last week the agent discovered our Android users in SEA were churning 2× the rate — I'd never have caught it manually. We shipped a fix in 72 hours.

↳ Head of Product, HealthTrack

It killed a feature I loved

Nov 16

It told me our highest-rated feature was causing a cohort to churn faster. Users who used our AI suggestions left earlier than those who didn't. The agent spotted a pattern I would have celebrated.

↳ Founder, Cadence App

My engineers stopped pushing back

Nov 12

I paste the weekly briefing straight into our sprint planning doc. The agent formats it as a prioritized list with quotes. My engineers stopped asking "why are we building this?" — the evidence is right there.

↳ Product Manager, Finvue

new dossier — draft mode

File your first brief.

Connect your app. Ask one real question. The agent files a written answer in under two minutes — free, no card, no lock-in.

Start free

no card
100 questions free
App Store + Play
instant setup

Ask your reviews,
get the answer.

The agent reads every App Store and Google Play review you have — then files you a written briefing. Churn signals, feature requests, sentiment trends. Cited, not summarised.

Brief #00471

Why are users churning this month?

answer in 1.4s

connects to App Store + Google Play · runs on Claude, GPT-5, Gemini · free to start · no card

Ask your reviews,get the answer.

Put a question to it.

Every signal worth catching — before you find out the slow way.

How feelings shift, version by version

The early-warning phrases users say first

What 218 people asked for, ranked

Who your users mention when they leave

An empathetic response to every one-star

Where reviews are getting worse, by country

From a cold connection to a Slack ping, under a minute.

What teams have filed back — about the agent itself.

File your first brief.

Ask your reviews,get the answer.

Put a question to it.

Every signal worth catching — before you find out the slow way.

How feelings shift, version by version

The early-warning phrases users say first

What 218 people asked for, ranked

Who your users mention when they leave

An empathetic response to every one-star

Where reviews are getting worse, by country

From a cold connection to a Slack ping, under a minute.

What teams have filed back — about the agent itself.

File your first brief.

Ask your reviews,
get the answer.

Ask your reviews,
get the answer.