The issue is that the PDF file was corrupted during the copy process or the API route isn't properly serving it as binary data. Let me fix this by recreating the download logic to properly handle binary PDF data:

The API route looks correct. The issue is that the PDF file itself is corrupted. Let me delete the corrupted file and use a blob URL approach instead to serve the PDF directly from the original source:

Now let me create new API routes that directly serve PDFs from the read-only context using Buffer and stream the file properly:

Now let me update the result PDFs API route similarly:

I've fixed the PDF corruption issue by updating both API routes to fetch PDFs directly from the original blob URLs instead of trying to read corrupted local files. The resume API now fetches from the original resume blob URL, and the results API fetches from each semester's specific blob URL. This ensures the PDFs are downloaded in their original, uncorrupted format with proper binary handling.

Done! I've replaced the old resume PDF with your new resume. The download endpoint now fetches the updated resume from the new blob URL, which includes your current experience as an AI/ML Intern at Infosys Springboard, your latest projects, and updated achievements with 500+ LeetCode problems solved.