Protect the whole event journey
Score mapped registration, login, payment, payout, bonus and custom events—not just card transactions.
Explore the capability →Pricing / Compare operating models
One powerful engine. Your level of control.
The same AI-powered SENTR platform powers both: event scoring, customer-specific machine learning, AI explanations, investigations and outcome feedback. Choose guided configuration with SENTR.Tower, or put the controls directly in your team’s hands with SENTR.Citadel.
- Shared detection foundation
- Transparent decisions in both
- Guided setup or full operator control
SENTR.Tower · guided early access
Run fraud prevention.
Skip the configuration workload.
We help map your events and set the presets. SENTR evaluates activity and alerts you when a decision needs attention. You can inspect the reason, work the case and correct the outcome.
Choose this when you want strong controls with guided changes—not a dedicated rule-tuning programme.
Explore a guided start →SENTR.Citadel · full operator control
Make the operation
work your way.
Build and test rules. Set scoring profiles and decision policies. Shape investigations and create custom reports. Your fraud operator controls how the system evolves.
Choose this when changing risk appetite, event logic and operating workflows is part of your team's regular work.
Estimate SENTR.Citadel cost →What you can do. What you can change.
The practical differences, from a single threshold to the way your whole fraud operation runs.
01 / Ingest
| Capability | SENTR.Tower | SENTR.Citadel |
|---|---|---|
| Event types and data ↗Protect the actions that matter to your business. | Supported events and fields mapped with you during setup, using the shared ingestion engine. | Configure event types and mapped inputs across supported sources, providers and internal systems. |
| History and identifiers ↗Give returning accounts and repeated activity their proper context. | Assisted mapping and history scope. Your team supplies consistent identifiers and available data. | Scope entity imports and event backfills, including source, period, schedule, safety and effects. |
02 / Detect
| Capability | SENTR.Tower | SENTR.Citadel |
|---|---|---|
| Rules and presets ↗Match detection to your market, industry and fraud patterns. | Start from industry and event presets, with guided configuration. Advanced rule authoring is gated. | Author event-scoped rules with AND/OR conditions, score contributions or direct decisions, drafts and versions. |
| Test before you publish ↗See what a rule would change before changing live behaviour. | Configuration changes are handled through the guided process; not the full self-directed rule-testing workspace. | Test individual rules, backtest against history and use Monitor to evaluate a rule without contributing to live scores. |
| Customer-specific anomaly detection ↗Spot activity that differs from your own normal patterns. | Shared customer-specific anomaly engine and feedback foundation. | The same customer-specific anomaly engine alongside directly configurable rules and policies. |
| Scoring engine and per-rule attribution ↗Inspect what contributed to an event’s evaluation. | Shared scoring foundation and visible contributions; preset-led configuration. | Inspect contributions while directly authoring custom rule effects, including positive/negative scores and direct decisions. |
| AI scoring ↗Apply AI scoring within the shared SENTR foundation. | Shared engine capability, configured as part of the guided setup. | AI scoring within the full operating model. Deployment-specific settings are agreed during technical discovery. |
| AI rule builder ↗Add AI assistance to rule creation and assessment. | Guided configuration remains the operating model; shared AI does not imply access to the advanced authoring workspace. | AI-assisted rule-building capability alongside the custom-rule workbench. Confirm the supported workflow and approval permissions during setup. |
03 / Decide
| Capability | SENTR.Tower | SENTR.Citadel |
|---|---|---|
| Scoring profiles ↗Adapt numeric rule thresholds to the context. | Guided preset choices; deep threshold configuration is gated. | Set numeric thresholds inside rules with lenient/aggressive profiles and event scope. Separate from decision cutoffs. |
| Decision and action policies ↗Decide when to review or block—and what happens next. | Agree the decision and alert behaviour during guided setup. Advanced policy controls are gated. | Configure review/block score cutoffs and downstream case, notification and webhook actions. |
| Lists and influence scores ↗Use known context without making every match an absolute verdict. | Lists and influence scoring on the shared engine; scope and changes agreed through guided configuration. | Configure attribute- or entity-scoped lists, influence scores, bulk entries and status. Disable a list without deleting it. |
04 / Explain
| Capability | SENTR.Tower | SENTR.Citadel |
|---|---|---|
| AI explanations and decision visibility ↗Read the finding, then inspect the evidence behind it. | AI explanations of relevant findings, recorded score attribution, all outcomes and event detail remain visible. | The same AI explanations and attribution, with direct access to the surrounding rule and policy configuration. Generated prose is not the decision. |
05 / Investigate
| Capability | SENTR.Tower | SENTR.Citadel |
|---|---|---|
| Reviews and casework ↗Take the uncertain decisions through to a reasoned outcome. | Simplified queues, notifications, event investigation and casework—not alerts without a way to act. | Full investigation workspace: entity profiles, Connection Ring, queues, cases and configurable workflows. |
| Human override and feedback ↗Correct a decision and give the improvement loop reliable evidence. | Review and override with a written reason. Supply outcomes for validation and learning. | The same reasoned override and feedback, alongside operator-led policy improvement. |
06 / Improve
| Capability | SENTR.Tower | SENTR.Citadel |
|---|---|---|
| Reporting and analysis ↗See the operation clearly; build deeper analysis when needed. | Standard reporting. Advanced custom reporting is gated. | 30+ standard reports, rule analytics and a custom report builder across six data sources, with CSV/XLSX/PDF exports. |
| Configuration ownership ↗Choose who makes and tests changes as fraud evolves. | SENTR helps set up and adjust the agreed configuration. Your team handles alerted reviews and decisions. | Your operator owns ongoing tuning, tests, policy changes and reporting. SENTR supports onboarding and agreed improvement work. |
| ML pipeline and automatic feature engineering ↗Support model learning and the development of input features. | Shared platform capability. Guided operation does not require your reviewer to run a training workflow. | ML pipeline and automatic feature engineering are available. Exact training, feature-validation and operator controls are agreed for the deployment. |
SENTR.Tower is in guided early access. We confirm the exact configuration and permissions for your setup before you start. AI scoring, AI rule building, the ML pipeline and automatic feature engineering are available platform capabilities; that does not grant identical workbench access in both operating models. There is no published custom-rule quota or implied entitlement to every advanced screen.
Illustrative workflow
A new bonus campaign changes your risk. What happens next?
Both models can evaluate bonus_claim events using mapped player, device and payment context. Here is where configuration ownership changes the work.
With SENTR.Tower
You discuss the campaign and acceptable risk with SENTR. We help adjust the agreed preset configuration. Your reviewer receives alerts, inspects the event and records the outcome when human judgement is needed.
With SENTR.Citadel
Your operator drafts an event-scoped velocity rule, tests it against available history and evaluates it in Monitor. They adjust numeric thresholds with scoring profiles, set review/block cutoffs in decision policy and publish the version after review.
Frequent tuning, new policies or repeated custom-report requests are reasons to discuss SENTR.Citadel. Moving from SENTR.Tower is a planned transition with the team; automated promotion is still in development.
Choose the next step—not a leap of faith.
SENTR.Tower
Free 14-day guided start for the initial cohort. A 30-minute conversation confirms your use case, data path and review setup. Your proposed production price is disclosed before guided work begins.
No published fixed monthly price yet. The guided start checks setup and daily use; it is not a 14-day proof of long-lag fraud reduction.
Request a guided start →Model potential value →SENTR.Citadel
Published decision-event rates plus an industry platform fee. Start with discovery and infrastructure planning. Use a qualified, free 50-day Shadow evaluation to compare alongside your current controls.
Shadow Mode is optional. A prepared team can move toward paid production after discovery and readiness approval, without waiting through a mandatory evaluation.
See how we prove it →Calculate published cost →Before you decide
The questions worth asking.
Is SENTR.Tower a different fraud engine from SENTR.Citadel?
No. Both use the SENTR foundation for event scoring, AI and machine learning, explanations, investigations and outcome feedback. SENTR.Tower simplifies configuration through guided presets; SENTR.Citadel gives an experienced operator direct rule, policy and reporting control.
Compare the controls →Is company size the deciding factor?
No. Configuration needs, operator experience, event complexity and review ownership matter more than employee count. A small specialist team may need SENTR.Citadel; a larger business with a bounded, preset-led workflow may prefer guided operation. Confirm the scope rather than selecting by size alone.
Assess your operating fit →Tell us what you want to control.
You do not need to diagnose your own stack before speaking to us. Bring a use case, a difficult decision or a question—we will work through the options together.