i’m pasha. i am a data analyst & software developer. i analyze data, tell visual stories, code apps and
websites. my stack includes python, javascript/typescript, node.js, react, django; pandas for analysis, plotly/bokeh and d3
for
data visualization.
my work in tech is driven by my reflections on tech. i work on projects that deal with the politics of
technology, its material/environmental costs, and resistance to the dominance of big platforms.
my work is sculpted by design, visual aesthetics, and storytelling.
open to collaboration. reach out if our visions intersect.
email: pasha.sem(at)proton.me
cables of resistance
2026 • project management&frontend
i took ownership of visual communications and the name of the ~900 participants conference connecting
activists,
researchers, hackers, artists, and others resisting big tech, as part of the public outreach team. i
collaborated with a designer to develop the visual identity and implemented frontend.
explore the website at cableresist.de.
semiconductor emissions explorer
2025 • data
freelance data project for interface’s global chip dynamics programme developing
the previous version of the project.
i conceptualized data-driven visual narrative and coded interactive d3.js
charts on emissions
from semiconductor
manufacturing. i was particularly
interested in how to translate complex context into coherent storytelling.
explore the interactive charts on the project's website.
the eye with no iris
2025 • reflections
a commission by the civic data lab, presented at silent green, berlin during the data festival.
in collaboration with abstraqt, we produced a video that explores the intertwined roots of
modern
statistics—and data science—with eugenics. the starting point for the video was our research that
galton, pearson, and fisher, referred to as the “fathers of modern statistics,” were also prominent
eugenicists.
watch the video in 3 parts on vimeo.
self-hosted music player
2025 • full-stack
private project born out of the need to share music between friends.
raspberry pi music server running gonic in docker, paired with a react/typescript frontend. custom
audio engine with queue management, nginx proxy, and ssh-based uploads
for multi-user access.
glu festival workshop
2025 • reflections
i talked about tech at an art and music festival glu.
my presentation started theoretically with ideas from langdon winner and donna haraway,
moved through current spotify’s leverage, and arrived at a practical section: presenting my
self-hosted music player and holding a workshop on creative approaches to tech together with abstraqt.
access full presentation on cryptpad.
semiconductor emission explorer
2024 • data
as part of my work at interface i analyzed data on
semiconductor emissions and built interactive charts with python plotly.
explore the interactive charts on the project's website.
tiktok auditing
2024 • data
as part of my work at interface i was auditing tiktok's recommendation algorithms: built a
pipeline to continuously
collect data from the tiktok ad library with python, analyzed collected
and survey data, ran statistical tests, and
produced visualizations and blog posts.
explore the whole tiktok auditing project.
fomometer
2024 • full-stack
a private project born out of my passion to curate berlin cultural events and share them with friends.
using plain html, css and vanilla js, i built a minimal directory with a simple table interface for
browsing and curation.
currently not updated as a more advanced v2 with django is under
construction but you can still
expore the site.
youtube's tree of complexity
2023 • data
as part of my work at interface/stiftung neue verantwortung i researched youtube’s platform
complexity, tracing recommendations, search, and news. i did exploratory
data analysis in python, ran statistical tests, and shaped the
publication story involving
scrolling interactive visualizations with d3.js. methods and findings
are documented in the
publication.
explore the scrolling visualizations on the publication's page.
germany's cybersecurity architecture
2023 • full-stack
as part of my work at interface i built a python pipeline to
turn excel data into interactive
bokeh visualizations for germany's cybersecurity policy research.
previously, every edition had to be
made manually as a pdf; the pipeline let non-technical users generate visuals automatically. built with
python, pandas, bokeh.
explore the interactive chart on the project's page.