3 threats need your attention
49 high-confidence scams blocked across email and voice analysis. Review queue below — playbooks are pre-loaded for each category.
Threats
Inbox
Scan email
IndicBERT + rule engine — 8 fraud categories.
Fraud Scanner
Voice analysis
Integrations
✓ READ ACCESSSubject, body, sender (in-memory)✓ STOREDMetadata + risk scores only✕ NEVERRaw email bodies, attachmentsTeam
Model health
Settings
Profile
Your personal details and language preference.
Sign-in & security
Manage how you access ScamShield AI.
Workspace
Used on invoices, exports and the DPDP grievance notice.
Billing & subscription
Trusted vendors
Suppliers whose domain ScamShield protects. Emails from look-alike domains will be flagged as vendor impersonation with risk score ≥ 90.
Team members
Invite teammates and set their access level. Admins can manage billing and integrations.
Per-category response
Block quarantines and alerts immediately · Flag adds a warning · Monitor logs silently · Mute ignores.
Trusted senders
Domains added here bypass rule-based blocking but still go through the ML model.
Blocked domains
Always-quarantine — emails from these domains skip analysis entirely.
Rule engine
Toggle individual rule modules. Disabling a category turns off only Component C; ML model still runs.
Email alerts
When should we email you about threats? Severity matches detection thresholds.
Quiet hours
Pause non-critical alerts during these hours. High-severity always goes through.
Org-wide security
Policies that apply to every member of your workspace.
Audit log
Recent admin actions — required for DPDP compliance.
API keys
Use the keys below to call /api/v1/analyse/email from your own systems.
Webhooks
Receive a signed POST callback when a threat is detected. Each delivery carries an X-ScamShield-Signature HMAC-SHA256 header.