Hi, I'm Soma!

I run a cat rescue and teach data journalism at Columbia's J-School.

In historical times I talked too much about food, co-founded a fake school, got quoted in Elle magazine, and ran a recipe website.

Upcoming Courses

The Automated Newsroom: Build AI Workflows That Work starts in November 2025 which seems like selling out because it costs money but oh lordy it's gonna be good.

Lede Program, a 10-week data journalism intensive. The... millionth? year starts Summer 2026.

Sites of interest

Practical AI for Investigative Journalism	a ~12 hour series about responsible use of AI in the newsroom
everything I know	about data journalism, anyway. mostly useful thanks to my perfect Python setup guide
Natural PDF	A friendly library for working with PDFs, built on top of pdfplumber.
aifaq.wtf	collected tweets and links about how weird and effed up the hypebeast world of AI is (featuring fish 🐟)
Normal AI	my favorite AI things that aren't ChatGPT
investigate.ai	baby-steps data science for journalists
Python's Not (Just) for Unicorns	an interactive introduction to programming in Python, for human beings and whoever else
Fancy GitHub	tips and tricks for making your GitHub profile shine
Data Viz Design Guide	a step-by-step design guide for data visualizations 🎉 📈
my YouTube channel	among other things, for some strange reason it's the most complete set of ai2html tutorials on the planet

Recent Talks

Protests, TikTok, and more 🎥: analyzing images and videos with AI	Dataharvest 2025
Browser puppetry: Playwright for dynamic website scraping	Dataharvest 2025
Everyday AI	IQ Media Conference
Second Draft: Everyday AI	Columbia Journalism Alumni Association
Utilizing AI in Journalism, A Workshop	New York Financial Writers Association
Fun little AI tricks for non-developers	CJI, 2025
Analyzing images and videos with AI	NICAR 2025
AI in the Newsroom	NICAR 2025
Passive Scraping for Social Media	NICAR 2025
Advanced Web Scraping - with AI flavor!	Lede Info Session, 2025
A Whole Bushel of AI Things	The Economist, 2025
Data Science Dojo	Tokyo 2024
BIRN Summer School (data/AI modules)	Balkan Investigative Reporting Network (BIRN) Summer School 2024
Turning Audio and Video into useable data	Abraji 2024
Investigating Documents with AI	Abraji 2024
Scrape state-level license and violations data with browser automation tools	IRE 2024
30 data cleaning tips in 60 minutes	IRE 2024
AI Everywhere: A Series of Human-Generated Talks	Media & Democracy Summit 2024
AI 202: Cooking with generative AI	NICAR 2024
How to make data viz with (almost) every tool	NICAR 2024
AI And ML: From Before And Beyond ChatGPT	NICAR 2024
Advanced Web Scraping with Playwright	Lede Program Info Session, 2024
Learn data journalism on your own!	Lede Program Info Session, 2023
Making and breaking AI in the newsroom	iMEdD Media Forum, 2023
AI + Journalism: In a world where robots can write like people, what's the role of the journalist?	Incoming CJS MS students, 2023
Reimagining the Humanities and Art with Generative AI	Fairfield University, 2023
Trusting AI in the newsroom: Hallucinations, bias, security and labor	ONA 2023
AI Tools from Before (and Beyond!) ChatGPT	ONA 2023
JournalismAI - Data and Technology	JournalismAI Academy for Small Newsrooms 2023
AI tool for Journalism Workshop	Abraji, 2023
Building and publishing with Quarto	C+J/European Data Journalism Conf 2023
How I convinced GPT to teach me about Hungarian folktales (without speaking a word of Hungarian)	MediaParty Chicago 2023
Building AI Editors	NICAR 2023
Practical Machine Learning for Everyone	NICAR 2023
Supercharge your coding workflow with ChatGPT and GitHub Copilot	NICAR 2023
An Introduction to ChatGPT for Educators (and maybe journalists)	Columbia Journalism School Faculty + Adjuncts, 2023

I love giving talks! Give me a shout at js4571@columbia.edu and we can make something happen.