Virtualized logic
Move selected JavaScript routines into a dedicated instruction set, adding distance between shipped code and its original intent.
Sentra Edge for Cloudflare Workers is now generally available
Bot mitigation for high-volume web apps
Stop scraping, credential attacks, transaction abuse, and fraud before they reach your app — without CAPTCHAs, slowdowns, or extra steps for real users.
Trusted in production by
Northwind
lumen/co
Atlas&Co
VELA
Halcyon
RIVERLINE
Mercato
stride.io
PARALLAX
Foundry
cinder.dev
QUARTZ
Mercer&Sons
Holberton
orbital/io
Pinecone
12ms
p95 verdict latency
0.012%
false-positive rate
4.3B
requests scored / day
1 PR
to integrate
Bot Defender
The only bot protection you need.
Every verdict at the edge — allowed, blocked, challenged, monitored — in one console. Drill into a session, see why it scored, and tune the rule that fired without leaving the page.
us-east-1 · last 24 hr
Bot Defender Overview
Monitor protected requests, enforcement outcomes, and risk concentration across your active sites.
Allowed
71.4%
43,240 sessions · low risk
Blocked
24.1%
3,847 sessions · 6 threat types
Challenged
2.5%
1,204 sessions · 87% resolved
Monitored
2.0%
962 sessions · shadow mode
Traffic Overview
Requests evaluated by Sentra over time, split by verdict.
Allowed
Challenged
Blocked
8K6K4K2K0
00:0003:0006:0009:0012:0015:0018:0021:00Now
Decision Breakdown
How the rule engine resolved traffic — allowed, challenged, or blocked.
-
Allowed43,24071.4%
-
Blocked3,84724.1%
-
Challenged1,2042.5%
-
Monitored9622.0%
Live Verdicts Live
| Session ID | IP address | Location | Risk | Action | Threat |
|---|---|---|---|---|---|
| sess_f9a2b341 | 91.108.4.1 | Moscow, RU | 94 | ● Block | Credential stuffing |
| sess_a1c93e02 | 203.0.113.45 | New York, US | 12 | ● Allow | — |
| sess_7d8f1290 | 45.33.32.156 | Fremont, US | 61 | ● Challenge | Suspicious timing |
| sess_e3190aff | 198.51.100.22 | Chicago, US | 88 | ● Block | Card testing |
| sess_b8d12904 | 185.220.101.4 | Amsterdam, NL | 91 | ● Block | API scraping |
Showing 1–5 of 482
…
Fraud Intelligence
See the pattern.
Stop the abuse.
A login, a payment, a refund. See how they connect. Bring account activity, device signals, and risk decisions into one investigation.
Activity explorer
Monitor account activity, risk signals, and policy decisions.
Events in view—Across the selected filters
Accounts—Distinct account identities
Blocked—Actions stopped by a policy
Needs review—Flagged for investigation
Event volume / Hourly
Recent activity
Latest 4 events · UTC| Time | Event | Account | Risk score | Decision | Location | Primary signal |
|---|
No activity matches these filters. Try another decision or reset your filters.
Shared device cluster3 accounts · 1 device · 2 network addressesSelect a node to inspect its evidence
Entity relationships
6 entities connected through device activity
Accounts in this cluster
Follow a connection into its session history042
Account 042
Overall risk94 / 100Combined session signals
Payment riskPatterns in payment activity
Access riskSign-in and device signals
Primary risk signals
Recent location
Approximate IP location
Device context
LAST SESSIONSession timeline
The sequence behind the decision
LATEST DECISION
Coordinated payment abuseCheckout activity · shared device · last 24 hoursComparison only — no policies are changed
Risk threshold
PROPOSED RULEBlock a checkout when:
Account risk is at least / 100
AND50 · more coverage95 · fewer matches
Device is shared by
3 or more accounts
3 or more accounts
The current policy uses a threshold of 90. Lower values include more sessions.
Decision comparison
Both rules evaluated against the same checkout activity
Current policyRisk ≥ 90
Proposed policyRisk ≥ 80
Differenceblocked checkouts
Current ruleProposed rule
These are rule matches, not confirmed fraud. Review the affected activity before changing enforcement.
Activity with a different outcome
| Time | Account | Risk | Current rule | Proposed rule | Shared-device signal |
|---|
No decisions change at this threshold.
Code Virtualizer
Your code. Harder to reverse.
Client-side code is exposed by design. We’re building a protection layer that makes sensitive JavaScript harder to inspect, modify, and reuse in automated attacks.
JAVASCRIPT PROTECTION · DEFAULT PROFILE
Code Virtualizer Overview
Upload JavaScript for obfuscation and configure your protection profile.
Obfuscate JavaScript
Select a file to prepare it for protection.
Drop your JavaScript file here
or browse from your computer
.js, .mjs, .cjs · Up to 10 MB · One file at a time
Upload your source file, then apply your protection profile.
No file selected
Protection starts with the routines you choose. Keep secrets, authorization, and pricing validation on the server.
The protection roadmap
Built in layers. Designed for your build.
Polymorphic builds
Vary transformed output between builds to make repeatable signatures and reusable analysis less straightforward.
Execution boundaries
Define intended domains and deployment windows for protected bundles, alongside server-enforced access controls.
Tamper detection
Check protected routines for unexpected changes and surface integrity signals to your application.
Runtime awareness
Look for debugging, instrumentation, and altered browser APIs that could change how sensitive logic runs.
Build fingerprints
Associate protected bundles with a release so teams can trace unexpected copies back to a build.
Response policies
Decide how your application should respond to integrity signals—from recording an event to requesting server verification.
Protection telemetry
Bring runtime signals into your monitoring workflow to investigate unusual behavior across releases.
Client-side protection adds friction; keep secrets, pricing validation, and authorization on the server.