01 Choose a model
Public02 Add your private data
BrowserReady.
Where Your Data Goes
Prepare in your browser
Validate ranges and normalize all inputs locally
Pack and encrypt the complete feature vector
No plaintext input upload
Compute on the server
Only ciphertext and evaluation keys are uploaded
Compute score under CKKS on the server
Encrypted features
All listed input values are encrypted. Model definitions are public; the secret key stays in your browser.
03 Result
Local decryptKey generation
Key upload
Compute
Encrypted result
–
Plaintext result
–
Latency
–
Supplier Upload
–
–
Admin Review
–
| ID | Version | Status | Size | Created | SHA256 | Action |
|---|
Encryption Params
log_n
–
slots
–
scale
–
q_chain_bits
–
security
–
–
Ciphertext Snapshot
Level / Scale
–
Moduli / logN
–
c0 head
–
c1 head
–
Ciphertext bytes (head)
–
Ciphertext hash (sha256)
–
Model & Timing
Model
–
Total latency
–