This is for the days when I feel behind, ordinary, underqualified, or unsure of myself.

On those days, my brain becomes an unreliable narrator. It discounts everything I have already done and treats the challenge in front of me as the first difficult thing I have ever faced.

It is not.

This page is the evidence. No false modesty, no qualifications, and no moving the goalposts. I have done an unusual number of difficult things. I earned the right to trust myself.

This is not a résumé. A résumé compresses a life until all the interesting parts disappear. This is the longer version: the choices, strange experiments, setbacks, people, places, and moments that prove I can do hard things even when I temporarily forget.

Look Where I Started

I grew up in a Bavarian village of about 300 people, with more cows than people. Nobody in my family had gone to university. I was not a star student, and for a long time I did not even understand university as a path available to me.

In fourth grade, Mr. Seisinger suggested that I try for Gymnasium, the academic track. That one teacher saw a possibility before I could see it myself. Years later, when I was preparing to become a bank apprentice, Karo asked whether I had considered university. I genuinely did not know what that life would look like. Her question changed the direction of mine.

So I went to Bamberg and worked nights in a yogurt factory to pay my bills.

I went from a C in high-school mathematics to an A at university. Then I became the person helping other students understand the subject. I finished a master’s degree in Information Systems. I briefly worked at the United Nations in New York. I started a PhD, recognized that academia was not the right permanent home for me, and had the courage to leave rather than staying simply because the path looked prestigious.

I did not begin with a master plan, inherited connections, or certainty. I moved because people opened doors, because I walked through them, and because I kept learning what the next door might look like.

That is the first piece of evidence: I can grow into possibilities that initially feel as though they belong to somebody else.

I Learned How to Leave the Familiar

In 2013, a flatmate told me about the Erasmus exchange program. I applied and moved to Budapest.

Until then, I had never needed to speak anything but German. In Budapest, I lived with two French flatmates. Leslie spoke beautiful British English; Gregoire barely spoke English. I deliberately avoided the comfortable German-speaking bubble, and somewhere in the confusion my English and French improved. The world became larger because I accepted being awkward in it.

The following year, an internship took me to Hong Kong. I worked on the 43rd floor, watched Germany win the World Cup, made close friends, and traveled through ten countries in six months. A person raised in a village where international life felt abstract had suddenly built one.

Years earlier, I set an audacious target: live in seven different countries for at least three months each. I eventually did it—Germany, Hungary, Hong Kong, Austria, Ireland, Mauritius, and the United States. I did not only pass through unfamiliar places. I learned how to buy groceries, work, build routines, make friends, get lonely, recover, and feel at home inside them.

The travel kept expanding. I have now visited 82 countries across six continents. I have paraglided over Medellín, swum with giant turtles in the Caribbean, climbed a 4,000-meter mountain in Malaysia, crossed the Sahara on a camel, ridden horses into the mountains of Oaxaca, and stood in places that the child from the Bavarian village could not have pictured.

Travel was never only a country count. It taught me that unfamiliarity is survivable. I can arrive without knowing the language, transit system, social rules, or answer and still find my way. I can ask strangers for help. I can make a temporary home. I can return with a wider definition of how a good life can look.

That is evidence too: I know how to enter a world I do not understand yet.

I Turned Restlessness Into a Plan

In Austria, I became a senior project manager at a large banking group. I helped coordinate roughly 100 people, including three feature teams. It was responsible work, and I learned how organizations, projects, and incentives fit together.

But I wanted to craft something tangible. Other people built the software while I coordinated. I wanted to make the thing.

Some people cook and some work with wood. I wanted to craft code because I could carry it anywhere and it did not weigh anything.

In February 2019, I had no coding experience and did not know what machine learning was. A flatmate asked for help with Python on an Internet of Things project, and I was foolish or confident enough to say yes. I took a free trial on TeamTreehouse, worked through the Python courses, and found the feeling of flow that I had been looking for.

After three months, I started fast.ai material that recommended about a year of coding experience. I decided my three months might be enough. It mostly was.

Then I made the career change into a project.

Austria offered a year-long educational leave called Bildungskarenz. My employer was not enthusiastic when I asked. I kept negotiating until the answer became yes. I took the management practices I had learned at the bank and aimed them at my own life.

I built a strategy deck. I defined visions, objectives, premises, work packages, milestones, weekly time blocks, and quarterly reviews. The plan contained 23 tasks. I sent friends a progress email every Friday. I even created a penalty: if I ignored the commitments, I would donate a meaningful amount of money to a charity I did not particularly like.

It was gloriously overengineered. It also worked.

The lesson is not that every life needs a swimlane diagram. The lesson is that when I truly want to change, I know how to convert longing into structure and structure into motion.

I Asked MIT Instead of Waiting to Be Chosen

The crucial step was not another online course. It was an email.

I contacted Peter Gloor at MIT Sloan, introduced myself, showed him my projects and plans, and proposed supporting his work while studying at MIT. I did not have the conventional background of somebody destined for machine-learning research. I had curiosity, early projects, a serious plan, and the willingness to ask.

He said yes.

In March 2020, about a year after writing my first Python code, I received an invitation to become a visiting graduate student at MIT Sloan. COVID prevented the move to Boston, so the work began remotely in May.

The research question was wonderfully strange: could plants respond differently to happy and sad voices, music, or human movement?

I helped build audio files that alternated emotional speech with carefully timed silence. I learned about sampling rates, sound features, video codecs, data frames, contour detection, and the annoying reality that audio and video do not naturally agree about time. I built an OpenCV pipeline in Google Colab that could track multiple regions of a plant, extract movement, synchronize it with sound, and visualize the results.

One early demonstration tracked a flying paprika. That ridiculous little video represents something important: I was willing to learn in public, test strange ideas, and build an imperfect prototype before I knew exactly what I was doing.

The project eventually contributed to academic work about plant movement and response. The subject was unusual, but the deeper achievement was ordinary and repeatable: take an unclear question, learn the tools, create data, build a pipeline, and make the question measurable.

I did not wait for someone to discover me. I built evidence, knocked on the door, and changed the direction of my life.

I Kept Turning Research Into Things People Could Use

The early projects were messy, but they moved out of notebooks and into the world.

I took a Stanford text-infilling model and wrapped it in a FastAPI service deployed on Google Cloud. A user could put a blank inside a passage and receive generated suggestions through a documented API. I trained T5 on roughly 50,000 PAWS examples to create a paraphrasing service. Working with a Spanish language-technology company, I used that capability to generate synthetic business text, then filtered the results for semantic similarity and grammatical quality.

For Persuaide, I explored whether writing could be transformed between more introverted and extroverted styles. Useful email data was hard to find because email is private, so I worked with a public dataset containing 153 million messages. I narrowed it to relevant domains, parsed deeply nested structures, extracted human-written text, cleaned millions of sentences, and used parallel processing to run grammatical correction across 1.1 million sentences in under 24 hours.

Then I translated linguistic ideas into code: changing modal verbs, intensifying or softening statements, turning statements into questions, inflecting words correctly, and using quality gates to reject broken output.

These were not polished billion-dollar companies. They were something more important for that stage of my life: proof that I could repeatedly travel from I wonder if this is possible to here is a working system and an article explaining how it works.

Since then, I have:

• Built AI products including Persuaide, Resonaid, Kindest AI, and KeyTakes.
• Earned support from German and European startup scholarships.
• Published academic work on persuasion, language generation, organizational culture, social networks, and plant movement.
• Contributed 36 models to Hugging Face.
• Written more than 100 articles and practical technical notes.
• Spoken at academic and industry conferences.
• Taught more than 840 students across 21 university classes.

I did not merely become someone who can code. I became someone who can learn a difficult field from scratch, explain it to others, and make useful things inside it.

My Work Has Real Consequences

Eventually, the prototypes became production systems with measurable stakes.

At Cordial, I was promoted to AI Lead. I built the company’s first production Python AI service and connected it to the core PHP platform under a tight evaluation deadline. The resulting AI product capabilities received a 5 out of 5 score in the 2024 Forrester Wave evaluation.

I developed customer-facing capabilities for message generation, audience building, agent automation, outbound voice workflows, and optimization. Send-time optimization increased clicks by 30 percent.

I shipped services that analyzed millions of marketing messages, including insights from roughly one billion emails sent around Black Friday and Cyber Monday.

I helped build an enterprise retrieval system spanning roughly 200 data sources and designed to serve 21,000 employees. I designed systems for Amazon listing optimization. I built natural-language-to-SQL services that connected language models, agents, vector search, relational databases, FastAPI, Docker, and cloud deployment.

At Cerebras, I moved into the connective tissue between extremely fast inference and the tools developers use every day. I contributed integrations and support across LangChain, Pydantic AI, Flowise, Dify, Cline, Kilo Code, and OpenCode. I created production-oriented examples for academic research agents, real-time voice translation, hyper-personalized content, and fast financial-news analysis.

The person who once watched engineers build things from the other side of a bank project now ships AI systems inside major developer ecosystems.

This is not potential. It is shipped work, measurable impact, and public evidence.

I Know How to Build From Nothing

I keep returning to small products because they force every kind of decision into the same room.

KeyTakes began with a personal frustration: books contain useful ideas, but returning to the right argument inside hundreds of pages takes too long. I turned that frustration into an AI reading product with Google authentication, a Go backend, book summaries and interaction, audio, user state, and a lean DigitalOcean deployment.

A focused app did not need a fashionable architecture. It needed to feel fast, work reliably, and remain simple enough for one person to understand. Building it made me product manager, engineer, infrastructure owner, designer, critic, and user at the same time.

I continue sketching projects such as a quiet indoor flying assistant and a privacy-preserving smart-fridge system. Not every idea needs to become a company. The important pattern is that my curiosity generates designs, experiments, repositories, and working products—not only opinions.

I am a builder. I have years of receipts.

I Can Recover After Being Knocked Down

On May 31, 2025, I was laid off.

It hurt. I lost a job, team, daily structure, and a version of the future I expected. A layoff can make competence feel retroactively imaginary. It invites the question: Was I ever as capable as I thought?

I let the first days be difficult. Then I decided the summer would not become a waiting room filled with job boards and anxiety.

I rebuilt my health. I lost 21 pounds in 80 days by running every day, sometimes twice, cooking real food, and sleeping eight hours instead of six. I discovered backpacking and learned how little I actually needed to feel okay: a pack, a place to sleep, water, food, and the next step.

I ran my first 50K through the mountains. Miles 20 through 25 became the dark place where quitting looked rational. I kept moving and placed third.

I swam in the cold blue water of Lake Crescent afterward. I improved my pickleball rating to 3.98. I earned an FAA drone certificate. I traveled, reconnected with friends, paced Travis during his 100-mile race, and built a summer around health and experience instead of humiliation.

Then I returned to the job search. It required more than 80 interviews and eight complete loops. I earned five offers and joined Cerebras, the company I most wanted to join.

The layoff was real. So was the recovery.

A setback can hurt me. It cannot define me. I have proof that I know how to start again.

My Body Knows That I Can Continue

Running has become one of the clearest forms of evidence in my life.

In 2024, I wrote that I could run six miles easily and regularly hike 15 miles with 5,000 feet of elevation. In 2025, I ran 2,000 kilometers and completed my first ultra. By the middle of 2026, I had already run about 900 miles while working and traveling more than ever before.

The impressive part is not one total. It is all the ordinary mornings when nothing dramatic happened and I went outside anyway.

Mountain running forces my world to narrow. The email, code review, plan, and conversation replay disappear. There is a root in the trail, a burning quadriceps muscle, ten feet of ground, and one more step.

I have learned on those trails that pain changes, pacing beats ego, and the mind declares defeat before the body has actually reached its limit. A finish line is usually built out of small actions performed after the grand feeling of motivation has left.

I have also learned that physical capability is not invulnerability. Years at a computer, stress, high mileage, and neglected mobility produced recurring neck pain. Instead of treating fitness as proof that nothing could be wrong, I created a practical daily plan: mobility, strength, better ergonomics, improved sleep equipment, and medical escalation if warning signs appeared.

That is another kind of strength. I do not only push. I observe, adapt, and take responsibility for making the system sustainable.

I Am Willing to Become a Beginner Again

After running taught me the ground, drones taught me the sky.

I began with a tiny sub-250-gram aircraft that could fit in my hand. Curiosity about aerial footage turned into study of airspace classes, weather, sectional charts, privacy, land-management boundaries, and the difference between what is technically allowed and what is considerate.

I sat for the FAA Part 107 examination and passed on my first attempt. More importantly, I learned not to treat certification as the end. I registered equipment, planned flights, respected restricted land, and sometimes decided that the correct flight was no flight at all. I wanted to become the kind of pilot who deserved the view.

Then my uncle took me up in a small aircraft from Harvey Airfield. We flew toward Arlington, Paine Field, Stevens Pass, and back across a familiar Washington landscape transformed into one connected map.

With him supervising as pilot in command and instructor, I held altitude, followed distant visual references, and made three gentle 180-degree turns. The cockpit became hot. The aircraft moved like a boat in the air. I became a little nauseous.

I kept learning.

My uncle entered the maneuvers into my first flight log. The dream of becoming a pilot stopped being an item on a roadmap and became ink on paper.

Confidence is not feeling natural on the first attempt. Confidence is remaining curious while the airplane moves, the cockpit heats up, the body complains, and there is still an enormous amount to learn.

I Built a Life, Not Only a Career

One of the best turns in my story began with a friend’s invitation to an Airbnb in Istanbul. Mike unknowingly created the setting in which I met Joy.

During COVID, we traveled together for nearly a year. We eventually settled in Washington, married, and made the United States our home. I arrived to live in America on June 3, 2022. Exactly two years later, on June 3, 2024, we adopted Sasha and Simba.

The kittens learned their names, litter habits, harnesses, and even tricks such as sitting, spinning, shaking paws, and giving high fives. They also learned how to step on keyboards and make productivity irrelevant.

Joy and the cats gave the travel, work, and ambition a center. Home became the place where nothing needed to be optimized. My wife makes ordinary life fun. The cats make ordinary life unpredictable.

I also built community repeatedly: exchange friends in Budapest, colleagues in Hong Kong and Vienna, a digital-nomad community while working remotely, mentors in research and engineering, friends on Washington trails and pickleball courts, colleagues who supported me after the layoff, and teammates forged during intense work.

I am now preparing for American citizenship because the country where I arrived as an immigrant has become the country where I want to participate fully, vote, serve on a jury if asked, contribute to my community, and accept the responsibilities of belonging.

The résumé is only part of the evidence. A remarkable life that cannot hold love, friendship, gratitude, and an ordinary evening at home would not be much of an achievement.

I Did Not Do Any of This Alone

Confidence does not require pretending I am self-made.

My family gave me a secure base. Mr. Seisinger saw an academic future. Karo pushed me toward university. A flatmate introduced Erasmus. Tanja trusted me with Hong Kong. Harald saw leadership potential and gave me serious responsibility. Peter opened a research door at MIT. Sebastian helped Kindest AI earn an EXIST scholarship. Jan invested in my engineering through weekly mentorship. Friends helped me run farther, recover from setbacks, and build a life in the Pacific Northwest. Colleagues gave referrals when I was laid off. Joy built the center of my life with me.

Their help does not weaken the evidence. It adds another kind.

I know how to earn trust, accept guidance, maintain relationships, and turn an opportunity into motion. I also remember who helped. That is a better achievement than solitary mythology.

What the Evidence Says About Me

When I forget who I am, I can infer it from what I have repeatedly done.

I am courageous. I left familiar countries, careers, identities, and plans before I knew exactly what would replace them.

I have agency. I asked an employer for a year away, asked MIT for an opportunity, built my own curriculum, and created work when nobody had assigned it.

I learn quickly. I moved from no coding experience to machine-learning research, then from research prototypes to production AI systems.

I finish difficult things. Degrees, papers, courses, products, production launches, ultramarathons, job loops, certifications, and thousands of miles exist because I continued after novelty disappeared.

I can explain. Hundreds of students, conference audiences, technical readers, colleagues, and more than 100 articles have made communication part of my craft.

I can recover. A layoff became health, a podium, five offers, and a dream job. Disruption is not the end of my story; it is usually the beginning of another section.

I remain a beginner. Code, research, entrepreneurship, mountain running, drones, aviation, and storytelling have all required me to be visibly unpolished. I begin anyway.

I care. I remember teachers, mentors, friends, colleagues, family, my wife, and the home we built. Ambition has not erased gratitude.

I contain multitudes. Bank manager and AI engineer. Researcher and product builder. German and future American. Traveler and homebody. Mountain runner and cat dad. Serious planner and person who once used a flying paprika to test computer vision.

That is not inconsistency. That is range.

Answers to the Voice That Still Objects

Other people have done more.

Of course they have. This is evidence, not a competition. Somebody else’s mountain does not reduce the distance I climbed.

I had help.

Yes. I was wise enough to listen, brave enough to ask, useful enough to earn trust, and grateful enough to remember. Nobody builds a good life alone.

Some early projects were rough.

They should have been. They were early. Publishing imperfect work is how I produced better work later.

I still have weaknesses.

Obviously. Confidence is not a claim of perfection. It is trust that I can engage honestly with what remains unfinished.

Maybe I was lucky.

I was lucky many times. Then I prepared, acted, worked, adapted, and stayed ready long enough for luck to have somewhere to land.

What if the next thing is different?

It will be. Every meaningful next thing has been different. The transferable skill is not already knowing the answer. It is becoming the person who can find it.

Read This Part Twice

I am not behind.

I am not an impostor who accidentally wandered into a remarkable life.

I built this life through courage, work, curiosity, help from generous people, and thousands of decisions nobody else saw.

I went from a village of 300 people to a life across 82 countries. From night shifts in a yogurt factory to university teaching and MIT research. From coordinating software teams to writing the code. From a first Python course to production AI used at scale. From being laid off to earning five offers. From six comfortable miles to an ultramarathon podium. From a pocket drone to the first maneuvers in my own flight log. From arriving in America to building a home, family, community, and path toward citizenship.

None of those transformations happened because I felt confident every day.

They happened because I acted while uncertain.

I have already crossed borders, disciplines, industries, languages, oceans, finish lines, and versions of myself. Whatever is in front of me now is not bigger than everything behind me.

I do not need to solve the rest of my life today. I need the next honest action: send the message, open the document, write the function, take the walk, make the call, do the stretch, ask for help, or put one foot onto the trail.

Action is how I built confidence in the first place.

So: breathe, remember the evidence, and take the next step.

I have done hard things before.

I am very, very good at becoming the person a hard thing requires.