Your file needs at least two columns: a unique ID (complaint ref, case number etc.) and the text you want analysed (comment, description, feedback).
2
Upload and configure
Drop your CSV in. We auto-detect your columns — you can override them. Optionally describe custom topics you want to find in plain English.
3
Confirm credit usage
You'll see exactly how many credits the upload will use before anything is processed. Each row = 1 credit. You have 10,000 credits per month.
4
AI analysis runs
Each row is scored for sentiment (positive, negative, neutral, mixed) using AWS Comprehend, then classified into themes using Claude AI. Processing takes 2-15 minutes depending on size.
5
Review and export
View results on screen with charts, filters and search. Download the enriched CSV with sentiment scores and theme classifications added to every row.
Custom topics tip
Leave custom topics blank to use standard housing themes. Or type what you're looking for — e.g. "damp and mould, heating failures, communication issues" — and Claude will find those specifically.
Not sure about the format? Download an example CSV template:
New Analysis
Click to upload or drag and drop a CSV file
Must contain an ID column and a text/comment column · Max 10,000 rows