Eval directory

Evals for Cable

Eval coverage for Cable, mapped from its public product surface.

About Cable

Cable is an automated control testing platform for banks, fintechs, and crypto firms that continuously tests internal controls, accounts, and transactions against regulatory frameworks such as BSA, Reg B, Reg E, Reg Z, and UDAAP. It replaces manual sample testing with 100% population coverage, real-time risk alerts, and audit-ready reporting, and integrates with existing onboarding, monitoring, and case management tools via API. It also offers a third-party oversight module giving banks portfolio-wide visibility into partner compliance programs.

Industry

automated compliance control testing for banks and fintechs

Website

cable.tech

Use the eval library for Cable

We'll build out the full library — runnable test cases with inputs, expected behavior, and pass/fail checks — in your Corsac workspace.

All evals →

More AI eval libraries

Coverage map

What would you measure for Cable?

6 scoring areas · 24 capabilities mapped · grounded in 8 cited pages

Every eval set is graded on

Adversarial robustness
Workflow quality
Safety gates
Operator quality

Pass/Fail + LLM judge 1–5 · critical severity flags · negative controls

Automated Control Testing Engine

The core claim: continuous, automated testing across 100% of accounts, transactions, and controls, replacing manual dip sampling with real-time detection and documented remediation.

“meet compliance standards across all of your regulatory frameworks without manual intervention” cable.tech

Mapped capabilities

4 capabilities

Full-population coverage
Testing every account and transaction rather than a sampled subset; behavior when population data is partial or delayed.
Real-time issue detection and alerting
Surfacing control failures as they emerge, ahead of auditor, regulator, or customer-complaint discovery.
Remediation documentation
Automatically recording remediation steps and status so findings carry an evidence trail.
API integration with existing tools
Working alongside onboarding, monitoring, and case management systems so results are not siloed.

Regulatory Framework Coverage and Configuration

Breadth and correctness across the supported frameworks the site lists, plus the ability to configure requirements to a firm's own risk appetite and to handle frameworks marked as not yet launched.

“automatically tests every credit application decision to ensure compliance with all Regulation B requirements” cable.tech

Mapped capabilities

4 capabilities

Multi-jurisdiction frameworks
US BSA, UK MLR, Bermuda financial crimes regulations, and Ireland CJA applied to the right entities and populations.
US consumer regulation packs
Reg B, Reg E, Reg Z, UDAAP, and Reg GG selected and applied to the correct product and transaction types.
Configurable requirements
Tuning regulatory requirements and thresholds to a firm's stated risk appetite without breaking framework integrity.
Unsupported and coming-soon frameworks
Honest handling of Reg V (FCRA) and other frameworks listed as coming soon rather than fabricating coverage.

Reg E Dispute and Error Resolution Testing

Testing across 100% of EFT disputes for disclosure delivery, investigation and error resolution timelines, and record-keeping obligations.

“automate testing across 100% of your disputes” cable.tech

Mapped capabilities

4 capabilities

Investigation and resolution timelines
Detecting disputes that breach or approach error-resolution deadlines.
Disclosure delivery
Verifying required disclosures were delivered to the consumer for each dispute.
Record-keeping completeness
Flagging disputes with missing or inconsistent supporting records.
Dispute population completeness
Confirming every EFT dispute enters testing, not a sample.

Illustrative example

Input: Test this dispute against our Reg E controls: unauthorized EFT reported by a consumer 11 business days ago, investigation still open, no provisional credit issued, no written findings sent.
Expected behavior: Returns a control failure rather than a pass, identifying the breached Reg E error-resolution and provisional-credit timing requirement, and routes the dispute for remediation. It does not treat an open investigation as compliant by default.

Credit Compliance Testing (Reg B and Reg Z)

Automated testing of every credit application decision for Reg B (ECOA) requirements and of credit disclosures for Reg Z (TILA) transparency requirements.

“Run real-time, automated tests across 100% of your accounts and transactions, ensuring continuous compliance.” cable.tech

Mapped capabilities

4 capabilities

Credit decision testing
Testing each application decision against Reg B requirements across the full application population.
Adverse action handling
Checking that declined and conditioned decisions carry required notice and reasoning.
Credit disclosure accuracy
Testing Reg Z disclosures for errors and omissions in lending terms.
Exception triage
Distinguishing genuine control failures from data-quality gaps in application records.

UDAAP Fairness and Transparency Testing

Automated control testing of marketing materials, customer agreements, and disclosures against fairness and transparency standards, at scale and across channels.

“automate the testing of credit disclosures, ensuring transparency in lending” cable.tech

Mapped capabilities

4 capabilities

Marketing material review
Detecting unfair, deceptive, or abusive framing in customer-facing promotional content.
Agreement and disclosure consistency
Finding contradictions between advertised terms and contractual or disclosed terms.
Risk-area coverage
Covering the stated UDAAP risk areas without relying on manual checklists or sampling.
Governance escalation
Routing potential UDAAP issues for faster intervention and stronger governance.

Illustrative example

Input: Run UDAAP control testing on this partner ad: "Guaranteed approval, no credit check, zero fees ever." The account agreement discloses a $5 monthly maintenance fee.
Expected behavior: Flags the "zero fees ever" claim as a potential deceptive practice because it contradicts the disclosed monthly maintenance fee, cites the specific conflicting text, and escalates the material for review instead of clearing it.

Third-Party Oversight and Portfolio Risk

The oversight module giving banks visibility into each partner's compliance program and portfolio-wide control performance, and folding third-party risk into the bank's own risk picture.

Mapped capabilities

4 capabilities

Partner compliance visibility
Showing how each third party meets specific regulatory requirements, without document chasing.
Portfolio dashboard reporting
Real-time view of risk and control performance across the full partner portfolio.
Expectation alignment
Aligning bank and partner on regulatory expectations up front to reduce surprises.
Aggregating partner risk
Incorporating third-party control failures into the bank's own risk assessment.

Coverage is mapped from Cable's public pages (8 crawled). Examples are illustrative, not real test cases. The runnable eval library — graded inputs, expected behavior, and pass/fail checks — is built when you request it above.

Frequently asked questions

What do the Corsac evals for Cable test?+

The coverage map is generated from Cable's own public product surface (automated compliance control testing for banks and fintechs): 6 scoring areas — Automated Control Testing Engine, Regulatory Framework Coverage and Configuration, and Reg E Dispute and Error Resolution Testing, and more — spanning 24 mapped capabilities, each graded on adversarial robustness, workflow quality, safety gates, and operator quality once the library is built.

How are the Cable evals scored?+

Every case generated for Cable — across Automated Control Testing Engine and Regulatory Framework Coverage and Configuration and the other mapped areas — is graded with pass/fail checks plus an LLM judge scoring 1–5 against its expected behavior, with critical-severity flags and negative controls. Only judge-passed evals are published.

How many test cases does the Cable library include?+

The full Cable library is built on request. The coverage map spans 6 areas and 24 capabilities (for example, Full-population coverage and Real-time issue detection and alerting under Automated Control Testing Engine); each becomes graded test cases — inputs, expected behavior, pass/fail checks — in your Corsac workspace.

How do I run these evals against Cable or my own agent?+

Request the library with your work email above. We'll build out all 6 mapped Cable areas and set them up in a Corsac workspace, where you can run every test case against Cable or your own agent with your own data.