
BuildU — AI Writing and Application Dashboard
Figures
Open full size
Open full size
Open full sizeOverview
BuildU is an AI-powered university application advisor. Users upload their personal statement, transcript, or other application documents; the app extracts the text via OCR, analyses it with GPT-4, and returns personalised feedback and strategy recommendations.
OCR Pipeline
Document uploads are processed client-side with Tesseract.js, which handles PDF, PNG, and JPG inputs and produces clean plain-text output. This approach avoids server-side file storage and keeps processing fast for the end user.
AI Analysis
The extracted text is sent to OpenAI GPT-4 with a structured prompt that asks the model to: (1) identify strengths and gaps in the applicant's profile, (2) suggest target universities matched to the profile, and (3) produce specific action items to strengthen the application. The response is streamed back to the UI using Next.js server actions.
Deployment
The app is deployed on Vercel with Next.js 15 App Router, using server-side rendering for the landing page and client components for the interactive upload and results flow. Edge functions handle the GPT-4 streaming response.