Find out what customers keep praising and complaining about.
Import reviews from a CSV or paste them in. Sentiment is broken down by topic, such as battery, price or support, with monthly trends that show when something changed and draft replies to unhappy customers.
The app itself, running inside this page. Anything you do here is saved in your browser.
Ask it to do things
Press Ctrl+K in the app (Cmd+K on a Mac) and describe what you want. The model chooses from the app's own functions, runs them in order and tells you what changed. Some examples:
| What is the biggest complaint and when did it start? | Reads the trends and answers with dates. |
| Compare battery feedback between the two products | Opens the comparison and summarizes it. |
| Draft replies to the five worst reviews | Writes replies into the queue. |
Instructions you save in the same box, such as "keep it short" or "I'm a beginner", are added to every AI request the app makes.
Features
- CSV import
- With auto column detection, paste mode, and a 146-review synthetic sample with real trends.
- Sentence-level lexicon sentiment
- With negation scope, intensifiers and contrastive 'but' weighting.
- Aspect extraction
- Diverging positive/neutral/negative bars (click to filter).
- Monthly rating vs text-sentiment trend lines
- And rating distribution.
- Distinctive-word clouds
- Via weighted log-odds ratio (Monroe, Colaresi & Quinn 2008).
- Suspicious-review detector
- Near-duplicates, bursts, generic 5, rating/text mismatch.
- AI executive brief
- Praise, complaints, requests, prioritized actions) and streaming reply drafts.
- Datasets
- Keep several imported or pasted review sets in the browser and switch between them; CSV import detects text, rating and date columns and rescales ratings out of 10 or 100.
- Trends page
- Monthly rating and text sentiment, a month-by-aspect heatmap of net sentiment, and change-point detection that reports when an aspect's complaints rose or fell.
- Phrases page
- Recurring two- and three-word phrases in negative and positive reviews, ranked by frequency and lift over the rest.
Runs in your browser
Parsing, sentiment, aspect mining, statistics, duplicate detection, charts. None of this needs a network connection or a model.
Uses a language model for
Executive summary from a stratified review sample + computed statistics. Customer-care reply drafting.
Notes
- Cost
- Free. Each account has a daily allowance of AI requests.
- Your data
- What you make stays in this browser on this device. Your account is only used for signing in and for the built-in AI. You can export a backup or clear everything from Settings.
- AI access
- Signed-in users get the built-in AI with nothing to set up. You can also use your own Groq or OpenRouter key under Settings.
- Source
- The code and tests are on GitHub under the MIT license. The tests also run in your browser.