My first company closed at 23. My current portfolio serves 1,500 people a day and carries the AI strategy of a global technology company. The path between those two points was not planned, but looking back, every stop trained me for the next one in ways I only see now.
TrendTalk, learning what money is for. At 21, alongside my engineering degree's final stretch, I raised $60K in seed funding and started a fashion aggregator app with a team of 12. We built good technology. Two years later we shut down, because we had spent like engineers: most of the budget into the product, too little into sales and distribution. By the time we understood where the money needed to go, there was no runway left to correct.
The lesson was expensive and permanent: where money goes matters as much as what gets built. Every product decision is a capital allocation decision. I have never forgotten it, and I've never needed to learn it twice.
Cognizant, learning discipline. After the shutdown, I took a Programming Analyst role in Cognizant's Data Science and Analytics practice, working for a US life-sciences client. Production data, a paying customer, deadlines that don't move. I automated their reporting operations and cut development time by around 40%.
It was my first sight of a pattern that would repeat for a decade: enterprises bleed time through manual, repetitive work, and the person who systematises it creates value that compounds. I also learned that I didn't want to service someone else's business. I wanted my own again.
The Funnel Hill Creamery, learning to run the whole machine. In 2017 I went back in, co-founding an American diner brand, this time with the unit economics worked out first. Over four years we grew to four company-owned outlets in Hyderabad, roughly $400K in annual revenue, a team of 70, and 250+ SKUs under standardised operations. We raised $200K on the strength of the first outlet to fund the rest.
Then COVID arrived. An incentive programme we built kept 93% of the team; a hygiene-first delivery campaign kept revenue moving; customer satisfaction rose from 80% to 92% through the industry's worst period. Four years of owning every lever at once, capital, operations, people, growth, with my own money carrying the risk. It is the closest education to product management that exists outside the job title.
The retool, learning better instruments. By year four I knew the business by feel, and I knew feel was the ceiling. Every major decision was intuition plus a spreadsheet, while the companies I admired decided with data and, increasingly, AI. So I made the strategic call: stepped away from a working business and moved to Germany for a Master in Data Analytics and Decision Science at RWTH Aachen, with applied roles alongside, explainable AI in pharma at Boehringer Ingelheim, analytics at utility scale at E.ON.
Those years taught me the second language, the engineering team's, to go with the boardroom's I already spoke.
Riedel, putting it all together. I joined Riedel Communications as a Data Scientist in Digital Transformation, and the role became the sum of everything before it. First, make the business visible: a BI layer across the ERP landscape that saved €500K+, master data driven from 70% error rate to 2%. Then digitise the processes: a pricelist cycle across seven departments cut from three months to under one. Then build the AI portfolio from zero: three GenAI products shipped, including a multi-agent IT helpdesk serving 1,500 employees with 40% ticket deflection, all with GDPR and EU AI Act governance built in.
Fifteen-plus digital products later, the pattern from Cognizant, the ownership from the founder years, and the instruments from the retool all run in the same role.
The thread. Every chapter was the same three moves at growing scale. Understand what the business actually needs. Translate it into something buildable. Orchestrate people and technology until it ships and pays for itself. The tools changed, spreadsheets, then SQL and Power BI, then multi-agent architectures. The job never did.