Summary
A demonstration analytics product for exploring revenue, targets, time-based changes, and revenue composition through an executive interface.
Project case study
- Demonstration project
- Synthetic data
- Web implementation
Public Revenue Intelligence Platform
A demonstration web analytics product for exploring synthetic revenue, targets, time-based changes, and revenue composition through an executive interface.
Private source codeVerifiable technical evidence is documented in this case study.
Overview
Problem
Fiscal datasets can be extensive and fragmented, making it difficult to identify trends, deviations, and relevant revenue sources.
My role
Indicator structuring, analytical scenario modeling, visualization development, and executive narrative design.
- Target attainment
- 97.2%
- Published demo / /api/revenue?year=2026, synthetic data / 13 Oct 2026
- Fiscal health score
- 99/100
- Published demo / /api/revenue?year=2026, concentration heuristic / 13 Oct 2026
- Monthly records in 2026
- 120
- Published demo / count from /api/revenue?year=2026 / 13 Oct 2026
Data and assumptions
The data is entirely synthetic and reproducible. A deterministic TypeScript generator creates monthly records with seasonality, trend, and controlled anomalies; the current source has no database, CSV ingestion, official API, or personal information.
An independent demonstration project with no institutional affiliation, developed to show how revenue indicators can support fiscal analysis and decision-making.
The project was inspired by knowledge gained from professional public-data analysis workflows without using internal systems, confidential data, or institutional materials.
- Volume
- 600 synthetic records
- Period
- January 2022 to December 2026
- Granularity
- One month per tax and fictional municipality
Data structure
Modeled entities
- Revenue fact: period, tax, category, estimated amount, collected amount, comparisons, and target status.
- Taxes: ISS, IPTU, ITBI, TLP, TLF, CIP, Active Debt, Fines and Interest, Property Revenue, and Other Revenue.
- Populated categories: Own-source Revenue, Property Revenue, and Other Revenue.
- Calendar: year, month, month name, and monthly key from 2022 through 2026.
- Municipality: one entirely fictional entity identified as mun-001.
Solution architecture
The flow separates data generation, access contract, calculations, APIs, and visualization to keep rules reproducible and allow a future real source without rewriting the interface.
Deterministic synthetic generator
Calendar, tax, seasonality, and seeded-noise inputs produce 600 TypeScript records identified as simulated-model-v1.
Data contract and provider
Domain types and RevenueDataProvider accept filters and return records and aggregates; the current implementation is exclusively an in-memory mock.
Calculations and analytical rules
Pure functions calculate totals, gap, attainment, changes, shares, accumulated values, forecast, score, and explainable insights.
Route Handlers and validation
Next.js APIs apply Zod-validated filters and return JSON in a consistent envelope.
Executive web dashboard
Server Components load the dataset, while client islands implement filters, the simulator, and Recharts visualizations in the responsive interface.
Indicators and business rules
The definitions below map to the application's TypeScript functions and methodology. Every result is calculated from synthetic data.
| Definition | Unit | Calculation rule | Use | Limitation |
|---|---|---|---|---|
| Estimated revenueSum of the reference values in the selected slice. | BRL | Sum of estimatedAmount. | Establish the comparison base for targets and projections. | The estimate is generated by the synthetic model, not by an actual budget law. |
| Collected revenueSum of the values simulated as collected in the selected slice. | BRL | Sum of collectedAmount. | Compare performance, composition, and time-based development. | The values do not represent official revenue or a real fiscal outcome. |
| Financial gapDifference between collected and estimated revenue. | BRL | Collected revenue − estimated revenue. | Identify positive or negative deviations from the reference. | The sign depends on two synthetic measures and does not express real budget impact. |
| Target attainmentPercentage collected relative to the estimated amount. | % | Collected revenue ÷ estimated revenue × 100. | Monitor adherence to the target in the filtered period. | It returns zero when the estimated base is zero and uses demonstration targets only. |
| Monthly changePercentage change in collected revenue between consecutive months. | % | (Current month − previous month) ÷ previous month × 100. | Observe acceleration, decline, and short-term seasonality. | The first point has no prior comparison and receives a zero change value. |
| Annual changePercentage change in aggregate collected revenue between consecutive years. | % | (Current year − previous year) ÷ previous year × 100. | Compare the growth trend across the five-year horizon. | The series includes the artificial trend defined by the generator. |
| Tax shareEach tax's share of total collected revenue in the selected slice. | % | Tax revenue ÷ total collected revenue × 100. | Analyze concentration and dependency in the simulated fiscal composition. | The weight of each tax was defined for the demonstration scenario. |
| Annual forecastClosing estimate based on the collection pace through the cutoff month. | BRL and % | Observed monthly average × 12, compared with the annual target. | Signal healthy, attention, or risk closing scenarios. | Heuristic forecast with a default cutoff at month 9 and no statistical model. |
| Revenue Health ScoreComposite score summarizing the health of the demonstration revenue slice. | 0 to 100 | Combines attainment, trend, concentration, below-target months, and anomalies with fixed weights. | Provide a consolidated executive reading of the selected slice. | Weights and bands are heuristic and were not calibrated with real fiscal data. |
Estimated revenue
Sum of the reference values in the selected slice.
- Unit
- BRL
- Calculation rule
- Sum of estimatedAmount.
- Use
- Establish the comparison base for targets and projections.
- Limitation
- The estimate is generated by the synthetic model, not by an actual budget law.
Collected revenue
Sum of the values simulated as collected in the selected slice.
- Unit
- BRL
- Calculation rule
- Sum of collectedAmount.
- Use
- Compare performance, composition, and time-based development.
- Limitation
- The values do not represent official revenue or a real fiscal outcome.
Financial gap
Difference between collected and estimated revenue.
- Unit
- BRL
- Calculation rule
- Collected revenue − estimated revenue.
- Use
- Identify positive or negative deviations from the reference.
- Limitation
- The sign depends on two synthetic measures and does not express real budget impact.
Target attainment
Percentage collected relative to the estimated amount.
- Unit
- %
- Calculation rule
- Collected revenue ÷ estimated revenue × 100.
- Use
- Monitor adherence to the target in the filtered period.
- Limitation
- It returns zero when the estimated base is zero and uses demonstration targets only.
Monthly change
Percentage change in collected revenue between consecutive months.
- Unit
- %
- Calculation rule
- (Current month − previous month) ÷ previous month × 100.
- Use
- Observe acceleration, decline, and short-term seasonality.
- Limitation
- The first point has no prior comparison and receives a zero change value.
Annual change
Percentage change in aggregate collected revenue between consecutive years.
- Unit
- %
- Calculation rule
- (Current year − previous year) ÷ previous year × 100.
- Use
- Compare the growth trend across the five-year horizon.
- Limitation
- The series includes the artificial trend defined by the generator.
Tax share
Each tax's share of total collected revenue in the selected slice.
- Unit
- %
- Calculation rule
- Tax revenue ÷ total collected revenue × 100.
- Use
- Analyze concentration and dependency in the simulated fiscal composition.
- Limitation
- The weight of each tax was defined for the demonstration scenario.
Annual forecast
Closing estimate based on the collection pace through the cutoff month.
- Unit
- BRL and %
- Calculation rule
- Observed monthly average × 12, compared with the annual target.
- Use
- Signal healthy, attention, or risk closing scenarios.
- Limitation
- Heuristic forecast with a default cutoff at month 9 and no statistical model.
Revenue Health Score
Composite score summarizing the health of the demonstration revenue slice.
- Unit
- 0 to 100
- Calculation rule
- Combines attainment, trend, concentration, below-target months, and anomalies with fixed weights.
- Use
- Provide a consolidated executive reading of the selected slice.
- Limitation
- Weights and bands are heuristic and were not calibrated with real fiscal data.
Technical decisions
Deterministic synthetic fiscal modeling
- Reason
- Create a reproducible series with trend, seasonality, and anomalies without exposing institutional data.
- Consequence / trade-off
- Realism depends on artificial parameters and cannot support conclusions about a real municipality.
Separation of data, calculations, and visualization
- Reason
- Reuse the same rules across pages, APIs, charts, tables, insights, and the simulator.
- Consequence / trade-off
- The architecture introduces contracts that must remain synchronized.
Responsive executive reading
- Reason
- Keep KPIs, filters, charts, and the table usable on desktop and smaller devices.
- Consequence / trade-off
- Analytical density must be distributed across blocks and viewport-specific representations.
Additional technical decisions
Server-first with client islands
- Reason
- Load data on the server and limit interactive JavaScript to filters, charts, insights, and the simulator.
- Consequence / trade-off
- Server and client boundaries require serializable contracts.
Replaceable provider
- Reason
- Avoid coupling the interface directly to the mock generator and prepare for a future authorized source.
- Consequence / trade-off
- The abstraction does not add persistence: the current implementation remains in memory.
Explainable insights and forecast
- Reason
- Centralize alerts, evidence, and recommendations in testable TypeScript rules.
- Consequence / trade-off
- Fixed rules and weights simplify auditing but do not replace statistical or fiscal validation.
Technical evidence
The application, endpoints, methodology, and local references are presented according to their actual access level.
Public application
Published application
Public landing page with a demonstration-project statement, synthetic KPIs, and access to the dashboard.
Open evidencePublic application
Executive dashboard
Public interface with filters, KPIs, score, forecast, charts, table, insights, and CSV export.
Open evidencePublic endpoint
Revenue API
Public endpoint with synthetic records, totals, ranking, categories, score, and forecast for the 2026 slice.
Open evidencePublic endpoint
Forecast API
Public endpoint with projected closing, gap, attainment, annual target, and scenario status.
Open evidencePublic documentation
Published methodology
Public page documenting fictional data, formulas, score, insights, and simulation.
Open evidenceImplementation references
Code reference
Synthetic data generator
Deterministic series from 2022 through 2026 with ten taxes, seasonality, and controlled anomalies.
Code referencesrc/data/mock-revenue.ts · src/data/dictionary.ts
Code reference
Indicator rules
Pure functions for totals, changes, shares, accumulation, forecast, and Revenue Health Score.
Code referencesrc/lib/calculations.ts · src/lib/forecast.ts
Private source code
Private repository
The source code is private; the published application and its APIs are the verifiable evidence.
Demonstration scenario indicators
- Simulated data
- BRL 248,587,119
- Collected revenue
- Sum of collectedAmount in the synthetic 2026 dataset.
- Simulated data
- 97.2%
- Target attainment
- Ratio between synthetic collected and estimated revenue in 2026.
- Simulated data
- 99/100
- Revenue Health Score
- Heuristic score calculated from the same demonstration slice.
Indicators calculated for the default 2026 slice from 120 synthetic records. They do not represent official revenue, a real fiscal outcome, or institutional performance.
Values reproduced by the published application and the /api/revenue?year=2026 and /api/forecast?year=2026 endpoints.
Demonstrated capabilities
- Fiscal indicator modeling.
- Time-based and comparative analysis.
- Executive narrative design.
- Visualization of synthetic public-sector data.
- Web analytics product development.
- Assumption and limitation documentation.
Limitations and next steps
Limitations
- Entirely synthetic data, municipality, targets, and outcomes.
- Independent project with no institutional affiliation or integration with government systems.
- No database, official refresh, or connected public source in the current implementation.
- Forecast, score, and insights based on demonstration heuristics.
- The repository is private, and the mockups do not reproduce the current API values.
Next evolutions
- Authorized integration with an official open-data source.
- Data-quality, provenance, and incremental-refresh validation.
- Expanded history and configurable deviation alerts.
- Authentication and access control for institutional scenarios.
- A real Power BI version documented separately from the web application.
- Observability and expanded automated testing.
Gallery
The images are product concepts and visual mockups. They do not represent a government platform or reproduce the values currently calculated by the published application.
Optional technical details
How to run
- Open the application published on Vercel.
- Open the dashboard to explore filters, KPIs, charts, rankings, and insights.
- Review the public methodology for formulas, score, and simulation rules.
- Use the public APIs to verify records and aggregates from the synthetic dataset.
Next project
Touchless Web Gesture Interface
Workspace controlled by hand gestures on the camera, with an adaptive frame pipeline and a custom gesture engine.
View full case