Cloud storage that reads, sorts,
and searches for you.
Upload documents once. Our AI extracts every word, categorizes each file, writes a short summary, and lets you find anything with natural-language search — even months later.
No credit card required · 100 MB free forever · Cancel any time
Everything you need to actually find your files
Six features that turn a plain cloud drive into an intelligent document workspace.
How it works
Three steps between "just an upload" and "instantly searchable knowledge".
Simple, transparent pricing
Start free. Upgrade when you outgrow the quota.
Your files stay yours
Every layer of the stack is hardened. We never touch your file bytes for upload or download — they stream browser ↔ S3 directly with short-lived presigned URLs.
- bcrypt password hashing (12 rounds) + password history (no reuse)
- TOTP two-factor with Google Authenticator + backup codes
- Email verification required before uploads / billing
- Account lockout after 5 failed login attempts
- Server-side session revocation & per-device sign-out
- Constant-time login (no email-enumeration timing leak)
- File magic-byte MIME check catches renamed malicious files
- CSP, HSTS, X-Frame-Options and 4 more security headers
- Rate limiting on every sensitive endpoint
- Every admin + auth event logged to an immutable audit trail
Frequently asked
Do you read my files?+
No — file bytes stream browser ↔ S3 directly via presigned URLs. Our servers only touch the file long enough to run OCR + summarization on your behalf, and the results (text summary, category, embedding) stay tied to your account.
What file types can I upload?+
Images (png, jpg, webp, gif, svg), PDFs and Office docs, plain-text formats (txt, md, csv, json), audio (mp3, wav) and video (mp4, webm). Each type has its own sensible size cap and a 1 GB absolute ceiling.
How does semantic search work?+
When you upload a file we compute a 1536-dim embedding from its OCR text + summary using OpenAI's text-embedding-3-small. Your search query is embedded the same way, and we cosine-match against every embedding via pgvector — no keyword match required.
Can I use this from a script or CI pipeline?+
Yes. Head to Account → API keys, generate a personal access token (aic_live_...), and pass it as Authorization: Bearer <token>. Every endpoint that accepts cookie auth also accepts a token.
What happens if I hit my storage quota?+
New uploads are blocked with a 413 response and a friendly upgrade prompt in the UI. Delete files to free space, or upgrade your plan in Billing — Stripe pro-rates the switch instantly.
Is this really a student project?+
Yes — final-year MCA project at Kalasalingam Academy, built by SRIRAM. Full stack: Next.js 14 App Router, PostgreSQL + pgvector, Prisma, AWS S3, OpenAI, Stripe. Source & docs in the repo.
Ready to stop losing files?
100 MB free — enough to try every feature. No card required.