Inspiration
Every business collects customer feedback, but most don’t know what to do with it. Valuable insights are buried in Google reviews, survey forms, and support logs—unread and unused. We wanted to fix that by making feedback instantly useful, without needing a dedicated analyst or hours of manual review.
What it does
Customer Feedback Analyzer is a web app that turns raw feedback (CSV or PDF) into clear, actionable insights. Users upload their files, and the app displays:
- Sentiment breakdown (positive, neutral, negative)
- Frequently mentioned keywords
- Grouped themes with user quotes
- Actionable recommendations
- Executive summary
All in a clean, interactive dashboard with no setup or technical expertise required.
How we built it
- Frontend: React (Vite) with Tailwind CSS for styling
- Backend: Node.js serverless functions on Bolt.new
- AI Engine: DeepSeek LLM
- File Handling: HTML5 File API and PDF.js for parsing
- Charts: Chart.js and word cloud libraries for visual insights
- UX: Responsive layout with accessibility support (WCAG 2.1)
- Deployment: Fully built and deployed in the Bolt.new environment
Challenges we ran into
- Structuring unstructured feedback into consistent themes and sentiment scores
- Ensuring fast, private, client-side file handling
- Designing an interface that was clean and accessible while still rich in functionality
Accomplishments that we're proud of
- Fully functional one-shot AI-powered product built without edits or follow-ups
- Real-time analysis and interactive dashboard experience
- Intuitive UX that makes AI insights understandable for non-technical users
- Multilingual support and complete browser-side privacy
What we learned
- Prompt engineering dramatically affects output quality
- A good UX is as important as the intelligence behind it
- You don’t need a large team or months of time to ship a real, usable AI product
- Bolt.new enabled us to go from idea to deployment in record time
What's next for Customer Feedback Analyzer
- Add custom prompt options and analysis fine-tuning
- Integrate direct survey platform imports (e.g. Typeform, Google Forms)
- Offer team workspaces and shareable feedback reports
- Launch to early testers and collect feedback to improve v2
Built With
- bolt.new
- chart.js
- cloud
- css
- deepseek
- file
- html5
- javascript
- library
- node.js
- papaparse
- pdf.js
- react
- tailwind
- typescript
- word
Log in or sign up for Devpost to join the conversation.