Venture Athlete & International Expansion Expert | Thriving as Generalist in a Specialized Economy

I’m Dominik, founder, athlete, & creator. A proud generalist. This is my story: 

For the last 7 years, I helped bootstrap one startup from 0 to 1M in revenue and built the international business from scratch for one of Europe’s fastest growing EdTech companies (50 -> 240 employees in my three years there).

All while competing at the highest level of the beautiful sport of Lacrosse, including an appearance for the German National Team at the 2022 World Games (the Olympics’ little brother).

While my professional achievements may seem more impressive on paper, representing Germany on an international stage is the feat that I’m most proud of.

I wasn’t very athletic up until my mid-twenties. Very little talent, but a lot of injuries to make up for it.

Until one day, I started working with a personal trainer. He helped me break down the fundamentals I needed to fix.

I worked hard, and I worked consistently. Showed up to every practice, every gym session. Did my recovery & mobility work daily.

4 years later, I made the national team roster.

The potential was in me. But it needed help unlocking.

Throughout my entire career, I’ve also helped humans unlock their potential, just in different ways:

- By teaching people in rural Rwanda and Brazilian homeless shelter how to make glasses for less than $1, and how to build a business around it
- By bringing unemployed nurses from Italy to Germany and teaching them German, opening a completely different career path for them
- By expanding the leading operating system for schools throughout Europe, so that education becomes more digital and thus accessible
- By writing about my experiences online and sharing them with the world

In late 2023, I made the decision to quit my cozy, well-paid, full-time job that fit my profile perfectly.

I'm on a mission to build my own portfolio career - and help others to unlock theirs. Join me on my journey towards building a portfolio of cashflow-positive, impact-driven businesses.

These businesses have one goal: help others unlock their potential. Just like my personal trainer did for me.

I wish that some of the best portfolio company owners turned creators had documented their journey from day 1, and not retrospectively.

So that’s what I’m doing.

I write for generalists like myself, who want to make a living in a specialized economy without specializing themselves. 

You can do more than one thing in your life. 
Let’s design that life together. 

Sign up for my newsletter here: www.dominiknitsch.com/newsletter

Dominik Nitsch

"Life is half over by 23" sounds like an insane take, but I think it has some substance to it. Here's why: 

When you're 5, each year represents 20% of your life. 
When you're 10, 10%. 
Etc.

With each passing year, the relative increase in experience slows down. I'm 32 now, so the current year represents ~3% of my entire life so far. And it's just gonna become less. 

So according to Sam, my life is already 75% over. Unacceptable. 

Here's what you can do about it: 

1/ Implement landmark events 
Some days, even some weeks, I ask myself: "wtf did I do today?" and can't really think of anything. It was just a normal workday with no specific things worth remembering. 

Chain too many of those days, and time passes by incredibly fast. 
But if you do something out of routine (like attending a sports event, traveling to a new place, meeting an interesting person, trying a new skill) – you'll remember that. And the more landmark events you have in a year, the longer the year will feel in retrospect. 

2/ Diversify, diversify, diversify 
If you just do one thing, it's much easier to get stuck in a rut. If you do several things really well, progress accumulates not just in your main job, but also in your side hustles. 

Subjectively, this makes life feel longer. You look back on a year and think: "damn, I did all this in three different disciplines? that's cool." 

3/ Engineer anti-routines into your life
I'm all for having a routine; the best work gets done in a familiar setting. 

But as with everything in life: "everything in moderation, including moderation". If you always stick to your routine, life passes by fast. So build anti-routines into it. These could be: 
→ setting an objective to travel to 2 new countries every year
→ consciously breaking your normal daily routine 1-2 days per week
→ meeting one new person every week 

Novelty makes life longer. 

That being said, Sam has a point. Life's short. Don't waste it doing things that mean nothing to you.

Founder @ Generalyst | building talent infrastructure for early-stage startups | 1,245 generalists read my newsletter to future-proof their career.

No alternative text description for this image

Specialization is the most overrated career advice of the last decade.

“Pick a niche.”
“Go deep.”
“Become the expert in one thing.”

What sounds great in theory, doesn’t work in practice anymore. 

It locks 20-somethings in boxes they shouldn’t lock themselves into yet. Most people in their 30s don’t know what they want. How do you expect high school and university grads to know that? 

The best way to find out what you want is to sample different things. 

This allows you to:
→ build a broad foundation
→ find out what you’re actually world-class at
→ get pattern recognition 

Knowing what you’re uniquely good at via trial and error is the ultimate way of “sharpening the axe”. 

Yeah, you might be falling behind compared to someone embarking on a medical or law career. 

But down the line, you’ll have a life that works exactly for you. 

Maybe it’s one job.
Maybe it’s a portfolio of jobs, turning into a Portfolio Career.

The possibilities are endless, but just picking one thing at 18 and sticking with it for the rest of your life sounds like a pretty bad strategy if you ask me. 

PS: I wrote a few articles about building a Portfolio Career. If that sounds interesting to you, you can download my portfolio career playbook here: https://lnkd.in/dGZaPsrv

Founder's Associates exist on a spectrum – from being purely operational to being a proper sparring partner. The breakdown: 

Want your own Founder's Associate? That's our bread and butter at Generalyst. We work with A+ generalist talents of all the three levels listed here. Faster than your own recruitment process, cheaper than your headhunter. 

PS: Want me to do one for Chief of Staff, too? Comment below.

Simon Veithen and I sat down on Friday to discuss: 

→ what the hell a generalist is, and why it's the best way to prepare for the future
→ how you can acquire meta skills 
→ why job safety is an illusion
→ how to achieve non-linear results 

... and much more. 

Listen to the full 1h episode here: https://lnkd.in/dNdxH_9s

How to identify high-performing organizations: 

"The single best indication of a high-performing culture is the prevalence of former entrepreneurs. It usually portends a high-agency organization equipped with people mentally and emotionally prepared for a startup journey. Figure out what percentage of a company’s employee base previously launched businesses of their own, and try to understand why they’ve decided to forgo their independence to contribute to a new mission." – Mario Gabriele

It takes a lot as a founder to subordinate yourself again (been there, done that). When an organization manages to attract these kind of people, there's usually something special going on. 

If you're a former founder and wanna join such an organization – there's a few in the Generalyst client portfolio :)

Carmine Paolino

Chat with Work

Founder and CEO

Freshflow

We’re reimagining the food supply chain starting with grocery retailers. Freshflow is a new way to manage fresh products in supermarkets. We do accurate order forecasting and recommendations that massively reduce food waste while still keeping shelves full. Doing that, we generate millions in savings for these businesses.

Today, more than 40% of all food produced globally is wasted. Besides the economic losses of $1.2 trillion/year that it causes, food waste is also the world’s third-largest emitter of greenhouse gases, right behind China and the US. There’s waste before the food even leaves a farm; it is then shipped, processed and stored, only to remain unsold in a grocery store. A majority of this waste is entirely avoidable, were it not for the antiquated and inefficient food supply chain.

Traditionally, store managers would have to think about weather conditions, external events, holidays, typical demand patterns, etc. when making an order decision. We do that with Machine Learning.
We’re working with some of the largest retailers in Europe and recently closed a round of funding from a top investor.

Are you ready to shape our success story?

Co-Founder and CTO

Entrepreneur First

Selected as part of a group of 50 to found and build a high-growth technology startup. The programme is backed by Reid Hoffman (founder of LinkedIn), the founders of DeepMind and PayPal, and some of the top investors (Founders Fund, Greylock Partners, Lakestar, etc.) in the world.

Founder in Residence

The Alan Turing Institute

Technical Mentor, Data Science for Social Good

OLX Group

End-to-end Data Science workflow (problem specification, requirements gathering, data gathering and cleaning using Redshift/Presto SQL and Pandas, modeling, coding of the services and pipelines in Python + asyncio and SQS/Kinesis/HTTP, productionization with Kubernetes and AWS, monitoring with Prometheus and Grafana, documentation with Confluence and Sphinx) of the following models:
* Image quality assessment model (MXNet)
* Image sharpness assessment model (OpenCV)
* Interpretable listing quality model (XGBoost + SHAP)

Spoke at GHOSTDay: Applied Machine Learning Conference 2019 about Scaling Machine Learning System up to billions of predictions per day.

Interviewed candidates for data scientist positions.

Wrote patches to MXNet Model Server to export Prometheus metrics.

Senior Data Scientist

Prosus Classifieds' (OLX, Avito, letgo) moonshot team; designing and implementing the future of classifieds.

* Taught an AI course
* Conducted qualitative user research and analysis using service design principles
* Conducted quantitative research
* Conducted market research
* Supervised a Master student's thesis on Out of Distribution detection

Senior Data Scientist @ OLX^2

End-to-end Data Science workflow (problem specification, requirements gathering, data gathering and cleaning using Redshift/Presto SQL and Pandas, modeling, coding of the services and pipelines in Python + asyncio and SQS/Kinesis/HTTP, productionization with Kubernetes and AWS, monitoring with Prometheus and Grafana, documentation with Confluence and Sphinx) of the following models:
* Prediction of listings' popularity (Scikit-learn)
* Similar items recommendations with a hybrid collaborative filtering and content-based recommendation algorithm (LightFM + FAISS)
    * User-item recommendations with a hybrid collaborative filtering and content-based recommendation algorithm (LightFM)
* Multi-task image classification (TensorFlow)
* Image classifier with thousands of imbalanced classes (TensorFlow)
* Image quality assessment model (MXNet)
* Image sharpness assessment model (OpenCV)
* Interpretable listing quality model (XGBoost + SHAP)

Keynoted at IFIP WCC 2018 with a talk on Deep Fine-Grained Image Classification.
Spoke at PyData 2018 on Semi-supervised Learning with GANs

Supervised a student on her Master thesis on a novel Saliency detection algorithm.

Lead teams that won several hackathons with the following projects:
* Suggest improvements to listings using interpretable machine learning (Scikit-learn + ELI5)
* Semantic search using feature embeddings from a hybrid recommender system (LightFM + FAISS)
* Real-time chat moderation (MXNet, transformer models)
* Create a listing by snapping a picture (TensorFlow)

Created a cookiecutter data science project structure with pre-setup Makefile, CLI, Sphinx documentation, automatic coding style using yapf, autoflake and isort, etc.

Interviewed dozens of candidates for data scientist positions.

Wrote patches to the following open source projects:
* MXNet Model Server
* sqs-requeue
* aiohttp-prometheus
* async-kinesis-client
* kinesis-python
* google's cld3

Data Scientist

Capgemini

Project Leader, Lead Architect and Big Data Engineer: Sport Innovator, VU, HvA, KNZB (Royal Dutch Swimming Federation) and KNRB (Royal Dutch Rowing Federation) were looking for ways to improve performance of top-level athletes through clever use of data. I led the team technically, architected and engineered the Big Data solution based on Hadoop, Spark, HBase, NiFi, Knox and Ranger, planned and assigned tasks in a Scrum team, communicating with the Product Owner, Scrum Master, Project Manager, Infrastructure team, and Partners, developed the first demo version of the front-end in AngularJS and D3.js, deployed the solution on Amazon AWS.

Big Data Developer, Train the Trainer program successfully completed, giving me the possibility to train other developers in Big Data technologies.

Senior Data Scientist and Big Data Engineer

Owlin

Developed a topic clustering algorithm, developed data visualizations, developed a bayesian spam-filter, refactored entire software projects, designed the company's current OpenStack cluster, developed software developers productivity tools for python and go developers (available at http://github.com/owlin), deployed software with Ansible, developed the company's authorization, authentication and OAuth provider, advocated the use of testing, continuous integration, continuous deployment, and containers, and selected the company's current bug tracker based on the company's needs; all in a Scrum team with bi-weekly sprints.

Data Scientist / Software Engineer

Gentoo Linux

Administered GNU/Linux machines, configured servers, deployed web applications.

System Administrator - Italian Community

Studio di Ingegneria Solare

Developed a solar panel installation recommender system from idea development to requirements gathering to user interface to implementation of a prototype.

Web Developer (Intern)

My journey in technology began at age 5, writing scripts to launch games – a simple start that sparked a lifelong passion for building solutions through code.

This drive to create led me to become deeply involved in open source, serving as Vice President and Technical Officer of the Italian Gentoo Linux community during my teens. It carried me through Computer Science at the University of Bologna, where my research on distributed systems and large-scale social network analysis was published by Springer, and then to studying AI at Vrije Universiteit Amsterdam, where I worked with an alumnus of Geoffrey Hinton on combining deep learning with my love for music.

Over the past decade, I've helped scale AI solutions at companies like OLX and Capgemini, led research teams at the Alan Turing Institute, and contributed to open source projects including MXNet Model Server and Theano. My work has ranged from building recommendation systems serving millions of users to developing novel approaches for data processing and analysis.

Currently, I'm founding two AI companies:
• Chat with Work - Making workplace knowledge instantly accessible through AI
• Vettr - Reimagining how VCs evaluate startups

Previously, I co-founded Freshflow, an AI company helping supermarkets reduce food waste through intelligent forecasting and inventory management.

When I'm not building AI solutions, you'll find me producing electronic music, DJing house and disco in Berlin's clubs, or exploring the world through photography.

Technical expertise: Deep Learning, Machine Learning, Distributed Systems, Python, Full-stack Development, AI System Design

Languages: Python, C++, Ruby, JavaScript
Frameworks: PyTorch, TensorFlow, Ruby on Rails

Just wrote up a nicer post about RubyLLM 1.14 on my blog.

Also launched my newsletter so you can get my posts in your inbox.

You can find a form to subscribe at the end of each post. Some of you already got the first email already!

Open article: RubyLLM 1.14: From Zero to AI Chat App in Under Two Minutes by paolino.me

RubyLLM 1.14: From Zero to AI Chat App in Under Two Minutes

taplio

The Battle

Dominik Nitsch

Carmine Paolino

Followers

Posts

Engagements

Likes

Performance ScoreDominik Nitsch's LinkedIn Performance

Performance ScoreCarmine Paolino's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Dominik Nitsch

Carmine Paolino

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Dominik Nitsch's LinkedIn Performance

Performance ScoreCarmine Paolino's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreDominik Nitsch's LinkedIn Performance