Radixing Request access
Deterministic rules + Jev model intelligence

Know your traffic.
Route with intent.

Resolve known traffic with deterministic edge rules. Add bounded model intelligence when the signals become ambiguous.

Request early access See the decision flow
Live decisions
00:00:00 UTC
RequestSignalDecision
24,891 requests evaluated p95 6.2ms
IP / ASNGEONETWORK TYPEBOTDEVICEJEV MODEL
01 / Platform

One decision layer.
Every request in context.

Radixing turns raw request data into a clear, programmable decision before traffic reaches your destination.

01Observe

Intelligence

Enrich requests with geo, ASN, network type, proxy, VPN, datacenter, bot, device, and risk signals.

02Decide

Rules

Compose readable policies from request context, with predictable priority and explicit outcomes.

03Act

Routing

Allow, block, redirect, route, or rate limit traffic at the edge without changing application code.

04Learn

Analytics

Inspect request history, traffic composition, rule hits, and outcomes from one focused operational view.

Tech preview / Jev
Shadow intelligence

Rules resolve the known.
Jev examines the uncertain.

Radixing is integrating Jev as a bounded second-stage classifier for traffic that deterministic rules cannot confidently describe. It reads sanitized signals, returns typed probabilities, and never bypasses the policy engine.

01 / Static engineAuthoritative

Deterministic first

Known fingerprints and hard policy matches stay on the fast, predictable rule path.

02 / JevBounded

Typed second opinion

Only eligible, unresolved traffic becomes a small set of sanitized features for probabilistic classification.

03 / Policy mapperLocal control

Safe by design

Versioned local policy decides what supported classifications may do. The model cannot invent a route.

Model
typesafe/jev-1.13
Input
Sanitized features
Output
Typed probabilities
Failure mode
Keep existing route
Use case / Conditional delivery

Conditional delivery,
made auditable.

Radixing turns request classification into explicit, testable traffic policies. Identify datacenter IPs, VPNs, proxies, bots, crawlers, devices, locations, and campaign traffic before routing each request.

IP filtering
Classify residential, mobile, hosting, datacenter, proxy, VPN, and Tor traffic.
Bot detection
Separate people, search crawlers, automation, scrapers, and suspicious requests.
Conditional routing
Route by geo, device, network, referrer, session, campaign, or custom signals.
Decision logs
Keep every traffic and routing outcome explainable, testable, and reviewable.

Built for legitimate conditional delivery, traffic protection, localization, quality assurance, and campaign verification. The same policy is visible to operators and applied consistently at the edge.

02 / Decision flow

From unknown request
to intentional outcome.

01

Identify

Capture request, network, device, and campaign signals.

02

Classify

Resolve identity, network type, automation, and risk.

03

Decide

Evaluate ordered policies against the enriched request.

04

Act

Send the request to the right destination and record why.

Decision preview POLICY / mobile-campaign-us
Incoming requestREQ-8F31
Country
United States
Network
Residential
Device
Mobile / iOS
Bot score
02 / 100
Campaign
launch-search
country == "US"true
device.type == "mobile"true
network.datacenter == falsetrue
DECISIONROUTE → /us-mobile
03 / Open core + hosted edge

Run the platform.
Inspect the engine.

Radixing's deterministic decision core is open source. Review, test, and self-host the policy engine, or use the managed platform for traffic signals, edge execution, telemetry, and operational controls.

Open source coreMIT · TypeScript · v0.2.0
@radixing/traffic-engine

Versioned policies, explainable rule evaluation, bounded traffic evidence, and an optional Jev decision gateway.

View source on GitHub
  • 01Deterministic rules, explained
  • 02Optional bounded Jev gateway
  • 03Hosted signals, execution, and audit
POST/v1/intelligence
// Request
{
  "ip": "203.0.113.42",
  "user_agent": "Mozilla/5.0..."
}

// 200 · 6ms
{
  "country": "US",
  "network_type": "residential",
  "proxy": false,
  "bot": false,
  "risk_score": 2
}
radixing-api200 OK
Questions / Answers

Traffic routing,
clearly defined.

What is conditional delivery?

Conditional delivery identifies a request and serves or routes it using signals such as IP type, location, device, campaign source, or bot status. Radixing expresses those decisions as readable policies with logs.

Can Radixing verify campaign traffic?

Radixing provides campaign verification, traffic classification, and conditional routing infrastructure. Teams control their policies and destinations while every decision remains explicit and auditable.

How is Radixing different from a WAF?

A WAF primarily decides whether a request is dangerous. Radixing determines what the traffic is, which policy applies, and where that request should go.

Can I run the decision engine myself?

Yes. The MIT-licensed traffic engine contains the pure policy and evidence core. Your host supplies trusted signals and executes decisions; Radixing's managed platform adds ingress, enrichment, telemetry, and operations.

Private preview

Make every request
an informed decision.

Radixing is opening early access for teams building traffic-sensitive products.

Talk to us