Hi, I’m Tiago Adria Nunes - an R/Shiny Developer with 10+ years in IT, specializing in data-driven web applications, interactive dashboards, and process automation. I’ve led development teams across banking, healthcare, education, and insurance, and I bring the same analytical rigour to both technical and business problems.
Projects
IMDb Top 5000
R Shiny ShinyApps.io
Interactive dashboard showcasing the top 5000 movies from IMDb, sourced live from IMDb datasets. Filter by title, director, genre, year, rank, rating, and vote count.
- Filter by title, director, genre, year, rank, rating, and votes
- TV Shows edition: IMDb Top 5000 TV Shows - same filtering approach applied to series and TV content
Last.fm Global Trends
Python Shiny Last.fm API
Interactive dashboard connecting to the Last.fm API to explore global and country-level music trends in real time.
- Real-time Last.fm API integration via pylast
- Global and country-level trend drill-down
Spotify Search App
R Shiny Rhino Spotify API
Search for artists, explore their profiles, top tracks, and related artists via the Spotify API. Includes genre-based discovery and network visualization of artist relationships.
- Artist profiles with top tracks and related-artist network graph
- Genre-based discovery with popularity and follower metrics
- API caching via memoise; built with the enterprise-grade Rhino framework
Databricks Ingestion Lakehouse
Python PySpark Databricks Delta Lake Docker
End-to-end data ingestion pipeline using NYC TLC Yellow Taxi data, organised into a Bronze → Silver → Gold medallion architecture. Runs both locally via Docker (PySpark + Delta Lake) and on Databricks (Unity Catalog + SQL Warehouse).
- Bronze: raw Parquet ingestion into Delta tables with metadata
- Silver: cleaning, filtering, and enrichment with derived columns
- Gold: analytics-ready aggregations for trip volume, revenue, zone performance, and time patterns
- Live analytics report published via GitHub Pages
Airflow DBT DuckDB Pipeline
Python Airflow DBT DuckDB DuckLake Grafana Prometheus Docker
End-to-end data pipeline orchestrating dbt transformations via Astronomer Cosmos on Apache Airflow, with DuckDB + DuckLake as the analytical engine and open table format, backed by a full observability stack.
- Daily DAG: TPC-H data generation → dbt Cosmos task group (one Airflow task per model) → staging and mart layers
- DuckLake open table format with Parquet data files and PostgreSQL catalog
- Observability: ducklake-metrics exporter → OpenTelemetry Collector → Prometheus → Grafana dashboards
dbt + pg_duckdb
Python DBT DuckDB PostgreSQL Docker
Dockerized PostgreSQL with the pg_duckdb extension and a dbt project for analytics transformations on TPC-H benchmark data. Every query is routed through DuckDB’s vectorized columnar engine via an on-run-start hook.
- Staging → Mart → Snapshot layers with enforced model contracts and SCD2 history
- Incremental models, automated data quality tests, and lineage DAG
- Live dbt docs auto-published to GitHub Pages on every push to main
Skills
Languages & Frameworks
R Shiny Quarto Tidyverse ggplot2 SparkR Python PySpark Pandas NumPy Matplotlib SQL
Data & Analytics
Data Modeling ELT/ETL DBT DuckDB Delta Lake Power BI Tableau
Process & Delivery
Business Analysis BPMN UML Scrum Kanban Jira Confluence
Education
- 2023–2024 - Postgraduate in Software Engineering - Descomplica
- 2006–2009 - B.A. in Business Administration - Anhanguera
Certifications
- Google Advanced Data Analytics Professional Certificate - Coursera, 2023
- Data Science: Foundations using R - Coursera, 2023
- Google Data Analytics Professional Certificate - Coursera, 2022
- Software Product Management Specialization - Coursera, 2018