Projects
I designed and built #OtwartePaństwo, a civic-tech platform that
consolidates Poland's scattered official journals into a single,
searchable and citable source of the law in force.
In Poland, a statute becomes binding not when parliament adopts it but
when it is promulgated — under Art. 88 of the Constitution, entry into
force is conditional on publication. Since 1 January 2012 the official
journals are no longer printed: their electronic edition, issued through
dziennikiurzedowe.gov.pl
and related portals, is the sole authentic source of binding law, and the
date of online publication is the date on which an act legally exists.
That law is dispersed across dozens of journals — the
Dziennik Ustaw, Monitor Polski and numerous ministerial
and regional gazettes — each with its own format and numbering. I developed
the platform with the Media 3.0 team to aggregate and normalize these
publications into one full-text corpus that can be searched, quoted and
linked with precision.
The platform makes the mechanics of legal validity legible: promulgation
dates, journal positions and effective dates become structured, queryable
data rather than isolated PDFs.
I co-created ktorzadzi.pl, the first Polish application to map the network
of connections between politics and business, built entirely from official
public registers.
Links between public figures and companies are formally public, but spread
across separate databases that are difficult to combine. The app draws on
official sources — the National Court Register (KRS) and the Central
Register and Information on Business Activity (CEIDG) — then cross-references
and reconstructs the relationships: who served on which board, where and
when, and how individuals and institutions interconnect.
I built it with the Media 3.0 team, in cooperation with Spain's
Fundación Civio, adapting the methodology of Civio's
Quién manda (“Who governs”) — a map of power that set a European
benchmark for data-driven accountability journalism between 2014 and 2023.
The contribution is methodological: transforming disconnected public
registers into a navigable graph of influence, with every link traceable
back to its official source.
I designed and built jaturzadze.pl, an application that makes municipal
budgets legible by processing and visualizing local-government financial
data.
Local budgets are published, but as dense spreadsheets that few residents
can interpret. The app ingests this financial data and renders it as clear
visualizations of where public money comes from and where it is spent,
letting citizens explore their own municipality's finances.
I created it together with colleagues from SAP Lab Poland and the
Media 3.0 team; the prototype won 3rd place at Poland's first
government open-data hackathon (Ministry of Digital Affairs, 2016) and I
later revisited it in a 2025 edition.
Aplikacja Parlament / Sejm
I created the first application to process live data from the Polish Sejm,
turning parliamentary activity into an accessible, real-time feed.
Votes, sittings, motions and MPs' work were hard to follow as they
unfolded. I captured the Sejm's data stream and presented it as a readable,
continuously updated feed, developed in cooperation with the Chancellery of
the Sejm (Kancelaria Sejmu).
It anticipated what is now standard practice — parliamentary transparency
as a live, machine-readable data source rather than static minutes.
I made Śladami Wojaczka, an augmented-reality, audio-literary walk through
the poetry of Rafał Wojaczek (1945–1971), the Silesian poète maudit.
I anchored Wojaczek's poems to the real places he lived and wrote in,
combining augmented reality (AR), a spoken audio layer and location-based
storytelling, with text mining of his corpus to surface recurring motifs.
It also works as an act of cultural-heritage archiving — preserving and
re-presenting a body of work bound to a specific city and landscape.
Literary heritage as something to be walked through: the city becomes the
interface, and each poem appears where it belongs.
Finnegans Work in progress
I'm working on Finnegans Work, a research project mining James Joyce's
Finnegans Wake for pattern and meaning across its many languages.
Finnegans Wake is written in a dense, portmanteau idiom that
interweaves dozens of languages and resists linear reading. I apply text
mining and semantic analysis, together with experiments in neural networks
and natural-language processing, to detect structural patterns and to
decode the work's multilingual layers.
The project treats one of literature's most demanding texts as a
computational-linguistics problem — a study in what machine reading can and
cannot recover from radically layered language.