About me
Good at computers.
Fun at parties.
I'm Linas Valiukas — a self-taught engineer with a journalism degree, 15 years of building software, and a decade at MIT Media Lab. Now I help European businesses figure out what AI can actually do for them.
Photo of Linas
From newsrooms to neural networks
I studied journalism at Vilnius University. Wrote articles, did interviews, learned to ask good questions. But somewhere between writing stories and building the websites that published them, I realized I liked the building part more.
So I taught myself to code. No bootcamp, no CS degree. Just curiosity, Stack Overflow, and a lot of late nights. Turns out, that combination of asking the right questions (journalism) and knowing how to build things (engineering) is surprisingly useful.
By 2012, it got me to MIT.
A decade at MIT Media Lab
I joined MIT Media Lab in 2012 to work on Media Cloud — one of the world's largest open-source news analysis platforms, built in partnership with Harvard University. I stayed for ten years.
When I started, Media Cloud was a ~100,000-line Perl monolith. I led its transformation into a distributed Python system that could process millions of news articles across 20+ languages. We built NLP pipelines, word2vec models, automated transcription — the kind of work that's trendy now but was genuinely hard back then.
The team's work got noticed. Our research was cited by the New York Times and appeared in over 600 scholarly articles. We secured more than $30 million in funding from the Gates Foundation, Ford Foundation, and others. I published at ICWSM 2021 (an AAAI conference) and spoke at PostgreSQL conferences in Singapore and Silicon Valley.
Ten years of turning messy, large-scale data problems into systems that actually work. That's the background I bring to every client engagement.
WordPress.com and the art of saving money at scale
In 2022, I moved to Automattic — the company behind WordPress.com — as a Senior Software Engineer. Different scale, different problems, same approach: find what's broken, fix it, measure the result.
I cut $500,000 a year from their AWS bill. Made backup restores 26x faster. Inventoried a 2.08 petabyte Amazon S3 store — that's roughly 2 million gigabytes of data, organized and accounted for.
Big companies have big problems. Working on them taught me how to think about cost, reliability, and systems that can't afford downtime. Those lessons shape how I approach AI deployments today — even for businesses with just a handful of employees.
Then I built my own thing
In 2024, I founded aero.zip — a privacy-focused file transfer service. Hit 1,000 monthly active users within two months. No ads, no tracking, no selling user data. Just fast, private file transfers.
Building aero.zip reminded me what I love most: taking a real problem, shipping a clean solution, and watching people use it. It also deepened my obsession with privacy — the same obsession that drives how I approach AI deployments at Lobster Pack.
By the numbers
15+
years in software engineering
10
years at MIT Media Lab
$30M+
in team funding secured
600+
scholarly citations
$500K
AWS costs cut per year
26x
faster backup restores
2.08 PB
S3 store inventoried
20+
languages in NLP pipelines
What I work with
Why I started Lobster Pack
When open-source AI agents exploded in late 2025, I saw something I'd seen before at MIT: powerful technology that most businesses couldn't access because they didn't have the right people to set it up.
Big tech companies have whole AI teams. European SMBs don't. But the tools are here, they're open-source, and they're good enough to make a real difference. Someone just needs to configure them, deploy them securely, and make sure they actually fit into your daily work.
That's what I do. I'm the person who bridges the gap between what AI can do and what your business needs it to do. Based in Vilnius, working across Europe, and genuinely excited about this stuff.
Want to work together?
Book a free 30-minute call. I'll listen to what your business does, ask some questions, and give you an honest take on whether AI automation is worth exploring.
Book a free call