Maritime Financial Analytics

The maritime vessel financial analysis system provides comprehensive budget oversight through:

1. Multi-dimensional Analysis: Budget variance, expenditure tracking, and committed cost monitoring
2. Mathematical Precision: Accurate calculations with proper temporal adjustments
3. AI-Enhanced Insights: Intelligent variance analysis with natural language explanations
4. Interactive Reporting: Drill-down capabilities and configurable visualizations
5. Scalable Architecture: Production-ready deployment supporting large vessel fleets

​

Data sources we use

• Budget Allocations: Monthly budget data across expense categories (OPEX, Non-Budget, Dry-Dock, Pre-Delivery)
• Expense Transactions: Actual financial transactions with posting dates and account classifications
• Committed Costs: Purchase orders and future financial obligations
• Period Definitions: Reporting periods and operational day calculations
• Fund Receipts: Capital inflows and budget allocations

---

​

Financial modules assessed

The system comprises four specialized analysis modules:

• Budget Variance Analysis: Core financial variance calculations with fund status tracking
• Expenditure Overview: Detailed tabular reporting with interactive drill-down capabilities
• Committed Cost Analysis: Purchase order tracking and commitment monitoring
• AI-Powered Monthly Variance: Intelligent variance insights with GPT-based explanations

---

​

System Architecture

The SIYA financial analytics platform follows a comprehensive ETL architecture that processes data from multiple ERP sources through intelligent transformation and analysis layers:

​

Architecture Components

External Data Sources

• Multiple ERP systems providing vessel financial data
• Real-time and batch data feeds
• Standardized data formats and APIs

SIYA ETL Pipeline

• Data Ingestion Layer: Handles multiple data source connections and formats
• Transformation Engine: Applies mathematical calculations and business rules
• Validation & Deduplication: Ensures data quality and removes duplicates
• AI Processing Module: Leverages machine learning for intelligent insights

Storage & Analytics

• MongoDB Cluster: Scalable document database for financial data
• Analytics Engine: Real-time processing and calculation engine
• ML Models: AI models for variance prediction and anomaly detection

Output Systems

• MCP Integration: Model Context Protocol for seamless data access
• Financial Reports: Automated report generation and distribution
• Executive Dashboards: Interactive visualizations and KPIs
• Variance Alerts: Proactive notifications for budget deviations

---

​

Mathematical calculations

​

Prorata Budget Allocation

For time-sensitive expense categories (CREW WAGES, LUBE OIL CONSUMPTION, MANAGEMENT FEES), the system implements proportional budget allocation: $\text{Adjusted Amount} = \frac{\text{Budget Amount}}{\text{Standard Period}} \times \text{Actual Period}$ Where:

• Budget Amount: Original allocated budget for the category
• Standard Period: Standard time period (typically 30 days)
• Actual Period: Current elapsed days in the reporting period

​

Year-to-Date Budget Normalization

Budget amounts are adjusted for partial reporting periods using two methods: Standard YTD Calculation: $\text{Normalized Budget} = \frac{\text{Monthly Budget}}{\bar{d}} \times d_{current}$ Where $\bar{d} = 30.416667$ (average days per month: $\frac{365}{12}$) Period-Based YTD Calculation: $\text{Normalized Budget} = \frac{\text{Monthly Budget}}{\text{Period Days}} \times d_{current}$

​

Variance Analysis

The core variance calculations used across all modules: Absolute Variance: $V_{abs} = B - A$ Percentage Variance: $V_{\%} = \frac{B - A}{B} \times 100$ Committed Cost Variance: $V_{cc} = B - (A + C)$ $V_{cc\%} = \frac{B - (A + C)}{B} \times 100$ Where:

• $B$ = Budget amount
• $A$ = Actual expenses
• $C$ = Committed costs (future obligations)

​

Daily Rate Analysis

For operational efficiency assessment: $R_{daily} = \frac{A}{D_{ops}}$ $R_{cc} = \frac{A + C}{D_{ops}}$ $R_{budget} = \frac{B}{D_{ops} + 1}$ Where $D_{ops}$ = Number of operational days

​

Fund Balance Reconciliation

The system tracks fund status across multiple categories: $\text{Balance} = \sum_{i=1}^{n} F_i - \sum_{j=1}^{m} E_j$ Where:

• $F_i$ = Individual fund receipts
• $E_j$ = Individual expenses

Surplus/Deficit Classification: $$\text{Status} = \begin{cases} \text{Surplus} & \text{if Balance} > 0 \\ \text{Deficit} & \text{if Balance} < 0 \\ \text{Balanced} & \text{if Balance} = 0 \end{cases}$$

---

​

AI-Enhanced Analysis

​

Fuzzy String Matching

For expense consolidation and duplicate detection: $\text{Similarity Score} = \frac{\text{Matching Characters}}{\text{Total Characters}} \times 100$

​

Commitment Percentage Calculation

For committed cost analysis: $C_{\%} = \frac{C}{B_{OPEX}} \times 100$ Where $C$ = Total committed costs and $B_{OPEX}$ = OPEX budget allocation

---

​

What is displayed in each module

​

Budget Variance Analysis Module

• YTD Budget vs Actual: Comparison of year-to-date budget against actual expenses
• Fund Status: Surplus/deficit analysis across fund categories
• Daily Rate Metrics: Per-day OPEX calculations and budget comparisons
• Variance Percentages: Detailed percentage variance analysis

​

Expenditure Overview Module

• Interactive Tables: Drill-down capabilities with monthly expense breakdowns
• Category Analysis: Hierarchical expense categorization and sorting
• Period Comparisons: Month-over-month expense tracking
• Prorata Indicators: Identification of prorata-calculated expenses

​

Committed Cost Analysis Module

• Purchase Order Tracking: Detailed PO information with supplier data
• Temporal Analysis: Current vs previous month commitment comparison
• Budget Impact: Percentage of OPEX budget committed
• Category Breakdown: OPEX, Non-Budget, Dry-Dock, and Pre-Delivery commitments

​

AI-Powered Monthly Variance Module

• Intelligent Narratives: GPT-generated explanations of variance causes
• Top Expense Analysis: Automated identification of significant expenses
• Monthly Summaries: AI-powered monthly variance explanations
• Trend Analysis: Historical variance pattern recognition

---

​

Expense categorization

The system uses hierarchical expense grouping:

Group	Categories	Description
OPEX	CREW WAGES, CREW EXPENSES, VICTUALLING EXPENSES, STORES, SPARES, LUBE OIL CONSUMPTION, REPAIRS & MAINTENANCE, MANAGEMENT FEES, MISCELLANEOUS, ADMINISTRATIVE EXPENSES	Operational expenses
NB	INSURANCE, NON-BUDGETED EXPENSES, P&I/H&M EXPENSES, VOYAGE/CHARTERERS EXPENSES, CAPITAL EXPENDITURE, EXTRA ORDINARY ITEMS, VESSEL UPGRADING COSTS, LAY-UP COSTS	Non-budgeted items
DD	DRYDOCKING EXPENSES	Dry-docking specific expenses
PD	PRE-DELIVERY EXPENSES	Pre-delivery expenses

---

​

Performance indicators and color coding

Status	Variance Range	Color	Action Required
Within Budget	$\pm 5\%$	Green	Continue monitoring
Attention Required	$> \pm 5\%$ and $\leq \pm 10\%$	Yellow	Review and plan checks
Action Required	$> \pm 10\%$	Red	Immediate investigation needed

---

​

Interactive features

• Drill-down Tables: Click-through capability for detailed expense analysis
• Period Selection: Configurable time ranges for trend analysis
• Category Filtering: Focus on specific expense categories
• Export Options: Multiple format support (CSV, JSON, PDF reports)

---

​

AI insights generation

The system leverages OpenAI GPT-4 for intelligent analysis:

• Variance Explanations: Natural language explanations of budget deviations
• Trend Analysis: Identification of expense patterns and anomalies
• Actionable Recommendations: Specific suggestions for cost management
• Monthly Summaries: Automated generation of financial performance summaries

---

​

Data quality and validation

• Prorata Identification: Automatic flagging of prorata-calculated expenses
• Duplicate Detection: Fuzzy matching to identify and consolidate similar expenses
• Data Completeness: Validation of required fields and data integrity
• Period Alignment: Ensures proper temporal alignment of budget and expense data

---

​

Report generation

The system generates comprehensive reports including:

• Executive Summaries: High-level financial performance overview
• Detailed Variance Analysis: Category-wise budget vs actual comparisons
• Fund Status Reports: Multi-dimensional fund balance analysis
• Committed Cost Summaries: Future obligation tracking and impact analysis

---

​

Example calculation: Budget variance analysis

Consider a vessel with the following monthly data:

• Budget Amount: $100,000
• Actual Expenses: $115,000
• Committed Costs: $25,000
• Operational Days: 30

​

Variance Calculations

$V_{abs} = 100,000 - 115,000 = -15,000 \text{ USD}$ $V_{\%} = \frac{100,000 - 115,000}{100,000} \times 100 = -15\%$ $V_{cc} = 100,000 - (115,000 + 25,000) = -40,000 \text{ USD}$ $R_{daily} = \frac{115,000}{30} = 3,833 \text{ USD/day}$ Status: Red (exceeds ±10% threshold) - Immediate investigation required

---

​

Best practices for interpretation

• Consider Multiple Dimensions: Analyze budget variance alongside committed costs and fund status
• Temporal Context: Review trends over multiple periods rather than single-point analysis
• Category-Specific Analysis: Different expense categories have different variance tolerances
• Operational Context: Consider vessel operational status and market conditions
• Proactive Monitoring: Use AI insights for early identification of potential issues

---