Dashboard

2 Scenarios / 8

Scenarios

Configure allocation scenarios and understand data sources

Active Scenario

Software Development Default

Code changes, bug fixes, feature development, security patches, architecture decisions

Financial Services

Transaction approvals, risk assessments, compliance checks, fraud detection

Healthcare

Patient triage, treatment recommendations, diagnostic support, medical records

Manufacturing

Quality control, production scheduling, equipment maintenance, supply chain

How Scenarios Work

Scenarios provide context for allocation decisions. The selected scenario influences:

Default threshold values
Risk assessment weights
Compliance requirements
Audit documentation levels

Software Development Scenario

Data Sources

In a production environment, MCAT integrates with your existing tools to extract decision context automatically:

Ticketing Systems

Jira / Linear / GitHub Issues - Issue type, priority, story points, labels
Extracted features: complexity estimate, affected components, linked incidents

Version Control

GitHub / GitLab / Bitbucket - PR size, files changed, test coverage
Extracted features: code churn, author experience, review history

Security & Quality Tools

SonarQube / Snyk / Dependabot - Vulnerability severity, code smells
Extracted features: security score, technical debt, dependency risks

CI/CD & Monitoring

Jenkins / GitHub Actions / DataDog - Build status, deployment frequency
Extracted features: failure rate, rollback history, incident correlation

Decision Types

Typical software development decisions and their recommended allocation modes:

Decision Type	Typical Mode	Why
Code formatting	AI	Low risk, deterministic
Dependency updates	AI	Automated testing catches issues
Doc typo fixes	AI	No functional impact
New API endpoint	Hybrid	Needs design review
Bug fixes	Hybrid	Verify root cause
Refactoring	Hybrid	Behavioral preservation
Security patches	Human	High impact, compliance
Auth changes	Human	Security critical
Architecture	Human	Long-term impact, novel

How Jira Fields Map to MCAT Dimensions

MCAT Dimension	Jira/Ticketing Fields	Example Calculation
Uncertainty (U)	Story points, description clarity, acceptance criteria completeness	`U = 1 - (criteria_count / expected_criteria) * clarity_score`
Novelty (N)	Labels, components, similar resolved issues count	`N = 1 - min(similar_issues / 10, 1.0)`
Value Conflict (W)	Watchers count, comment threads, priority changes, linked blockers	`W = (watchers * 0.1) + (priority_changes * 0.2) + has_blockers`
Data Quality	Required fields filled, attachments, reproduction steps	`DQ = filled_fields / total_required_fields`
Confidence	Historical success rate for issue type, assignee track record	`C = type_success_rate * assignee_success_rate`

Financial Services Scenario

Data Sources

In a production environment, MCAT integrates with financial systems:

Transaction Systems

Core Banking / Payment Gateways - Transaction amount, frequency, location
Extracted features: deviation from patterns, velocity checks, amount thresholds

Risk & Compliance

AML/KYC Systems - Customer risk scores, watchlist matches
Extracted features: compliance flags, regulatory requirements, audit trails

Analytics & Fraud Detection

ML Models / Rules Engines - Fraud scores, anomaly detection
Extracted features: model confidence, rule triggers, historical patterns

Customer Data

CRM / Account Systems - Customer tenure, account history
Extracted features: relationship value, complaint history, product holdings

Decision Types

Typical financial services decisions and their recommended allocation modes:

Decision Type	Typical Mode	Why
Standard transaction	AI	Within normal patterns
Recurring payment	AI	Predictable, low risk
KYC refresh	AI	Routine verification
Unusual transaction	Hybrid	Needs context review
Large transfer	Hybrid	Amount threshold
New account opening	Hybrid	Identity verification
Fraud alert	Human	High stakes, legal impact
SAR filing	Human	Regulatory requirement
Account closure	Human	Customer impact, compliance

Healthcare Scenario

Data Sources

In a production environment, MCAT integrates with healthcare systems:

Electronic Health Records

EHR/EMR Systems - Patient history, vitals, medications
Extracted features: condition severity, comorbidities, allergies

Clinical Decision Support

CDSS / Diagnostic Tools - Differential diagnoses, treatment protocols
Extracted features: guideline compliance, drug interactions, risk scores

Imaging & Diagnostics

PACS / Lab Systems - Radiology reports, lab results
Extracted features: abnormality flags, trend analysis, critical values

Scheduling & Workflow

HIS / Scheduling Systems - Appointments, bed management
Extracted features: urgency levels, resource availability, wait times

Decision Types

Typical healthcare decisions and their recommended allocation modes:

Decision Type	Typical Mode	Why
Appointment scheduling	AI	Routine optimization
Rx refill (stable)	AI	Maintenance medication
Documentation assist	AI	Administrative task
Triage assessment	Hybrid	Needs clinical validation
Diagnostic support	Hybrid	AI suggests, MD confirms
Drug interaction check	Hybrid	Alert review required
Treatment decision	Human	Patient safety, ethics
Emergency protocol	Human	Critical, time-sensitive
End-of-life care	Human	Ethical, family involved

Manufacturing Scenario

Data Sources

In a production environment, MCAT integrates with manufacturing systems:

Industrial IoT & SCADA

Sensors / PLCs / SCADA - Temperature, pressure, vibration, speed
Extracted features: anomaly scores, trend deviations, threshold breaches

MES & ERP Systems

SAP / Oracle / Custom MES - Production orders, inventory, schedules
Extracted features: order priority, resource utilization, lead times

Quality Management

QMS / SPC Tools - Inspection results, defect rates, SPC charts
Extracted features: Cp/Cpk indices, control limit violations, trends

Maintenance Systems

CMMS / Predictive Maintenance - Work orders, asset health, failure predictions
Extracted features: RUL estimates, failure probability, maintenance history

Decision Types

Typical manufacturing decisions and their recommended allocation modes:

Decision Type	Typical Mode	Why
Parameter adjustment	AI	Within control limits
Routine inspection	AI	Standard QC check
Inventory reorder	AI	Below reorder point
Schedule change	Hybrid	Multi-factor impact
Preventive maintenance	Hybrid	Production trade-off
Quality deviation	Hybrid	Root cause analysis
Line stoppage	Human	High cost, safety
Supplier change	Human	Strategic, contractual
Product recall	Human	Legal, brand impact