I build analytics systems that turn messy operational inputs into clean data, trusted metrics, and decision-ready outputs.
My current focus is combining business intelligence, analytics engineering, and AI-enabled workflows — especially in data centers, infrastructure market intelligence, Power BI, SQL, and reusable Python analytics systems.
- Data center transaction readiness and infrastructure market intelligence
- Modular analytics systems in Python
- Power BI dashboards, semantic models, and reporting workflows
- SQL / DuckDB / PostgreSQL analytics
- Metric governance, reusable KPI layers, and trusted reporting logic
- AI-assisted analytics automation and workflow design
A modular analytics system that converts a structured client intake workbook into:
- Readiness KPIs
- Deterministic transaction-readiness recommendations
- Executive report HTML + PDF
- Dashboard HTML + PDF
- Power BI handoff exports
- Machine-readable artifact manifest
- Curated client delivery package
Core workflow:
Client Intake Workbook
↓
Readiness Workbook Builder
↓
Analytics Pipeline
↓
Metrics + Recommendations
↓
Reports + Dashboards + Power BI Exports
↓
Client Package
Why it matters:
The toolkit helps data center occupiers, developers, brokers, and investors understand whether a requirement or project is actually ready to transact — before moving deeper into external engagement, site selection, or investment review.
🔗 Project repo: financial-data-analytics
The flagship toolkit is built on a reusable modular stack:
| Component | Purpose |
|---|---|
| Intake Engine | Cleans and validates messy CSV / Excel inputs |
| Metrics Engine | Produces governed KPI outputs from YAML-defined metric logic |
| Report Engine | Generates Markdown, HTML, and PDF reporting artifacts |
| Analytics Store | Loads trusted outputs into DuckDB tables and views |
| Visuals Engine | Renders self-contained offline dashboards |
| Analytics Pipeline | Orchestrates the full workflow end to end |
| Readiness Workbook Builder | Converts multi-sheet intake workbooks into analytics-ready exports |
- Cold-start reproducible GitHub demo
- GitHub Actions CI with package tests and end-to-end readiness integration validation
- 300+ automated tests across the active system
- Client-facing reports, dashboards, BI exports, and delivery packaging
A Python CLI tool for turning messy CSV / TSV / Excel files into analytics-ready outputs.
Capabilities include:
- Automated cleaning and normalization
- Validation and profiling reports
- DuckDB loading
- Batch processing
- YAML pipeline configs
- HTML quality reporting
A config-driven KPI layer for converting clean data into validated, reusable metrics.
Capabilities include:
- Schema-driven normalization
- YAML metric definitions
- Configurable segment rollups
- Sum-before-divide KPI logic
- Long and wide metric outputs
- Metric dictionary generation
- Validation report export
Clean data → Trusted metrics → Visuals anywhere
My goal is to build analytics systems where:
- messy source files become clean structured data,
- KPI logic is reusable and auditable,
- dashboards and reports consume trusted outputs,
- and future AI agents can reason over structured metrics rather than raw spreadsheet chaos.
Analytics & BI: Power BI, DAX, dashboards, semantic models
Data & Querying: SQL, PostgreSQL, DuckDB
Programming: Python, Pandas, Polars, automation
Workflow & Delivery: Git, GitHub, CI, CLI systems, YAML-driven configuration
Domain Focus: Data centers, infrastructure intelligence, transaction readiness, financial analytics
LinkedIn: linkedin.com/in/mat-horobjowsky