"Your ERP has the answer. Your spreadsheets have the answer. Your SAP instance definitely has the answer. Nobody can find it - until now."
Every large company has a data problem. Not a shortage of data - quite the opposite. The data is everywhere: in Excel files emailed on Fridays, in SAP modules nobody fully understands, in ERP exports that require three consultants and a prayer to decipher. The insight is in there. Getting it out takes weeks.
Leopold von Waldthausen and Marvin Christopher Bornstein watched this happen for years inside Palantir, where they spent a combined 15+ years doing exactly that work - untangling enterprise data for some of Europe's largest organizations. At some point, they stopped asking "how do we make this faster?" and started asking "why does this require engineers at all?"
Zeit AI is their answer. Type a question. Get the insight. No data engineer, no SQL, no ticket queue. Their product, ZeitMind, turns whatever data your company runs on into a system that answers questions in plain English - and shows its work.
Y Combinator backed them in the Summer 2024 batch - a cohort that was 75% AI-focused, and the most competitive in YC's history. Zeit wasn't selected because of a clever pitch deck. They were selected because the problem is enormous and the founders have seen it from the inside.
From data to enterprise insights in just a few words.
- Zeit AI's entire product brief, which is also their entire pitchZeitMind connects to where your data already lives - then gets out of the way.
Plug in Excel, CSV, SAP, ERP, CRM, databases - wherever the data lives.
Type a question in plain English. No SQL, no formulas, no special syntax.
ZeitMind locates, joins, and analyzes the relevant data automatically.
Get charts, reports, and actionable insights. Results are repeatable and explainable.
The short answer: anything that currently requires a data analyst and a week of back-and-forth. The longer answer includes these real use cases from their target market.
That's the internal shorthand. The reality is a structured intelligence layer sitting between your organization's raw data and the decisions that data should be driving.
Ask questions the way you'd ask a colleague. ZeitMind finds, joins, and returns answers from your actual enterprise data - not a sample, not a demo dataset.
Insights don't stop at text. ZeitMind produces charts and visualizations on the fly, formatted for decks, reports, and executives who skim.
Recurring reports that used to take analysts days can run automatically. The same logic, on fresh data, every time - with auditable, explainable outputs.
Zeit AI engineers work alongside your business teams - not behind a support ticket queue. Measurable results within weeks, not quarters.
GDPR compliant. Your data doesn't train their models. Period. For enterprises that can't afford ambiguity here, that's not a feature - it's the baseline.
Configure rules that flag anomalies and trigger actions automatically. The system doesn't just answer questions - it spots the questions you forgot to ask.
They didn't read about enterprise data chaos in a case study. They delivered software into it, for years, across industries, for companies that run on SAP and Excel and willpower.
Computer Science at Oxford. Then 3.5 years at Palantir, leading enterprise customers across Germany, Austria, and Switzerland. Published research on process mining and cryptography. And before any of that - founded a company called weview at age 18 and sold it.
The kind of founder who started selling before he started studying.
Writing software for over 18 years. Grew up near Berlin, spent five years in London, then five years at Palantir working across chemicals, railways, media, utilities, insurance, and medical equipment. Built an AI rule-engine that generated €50M+ in novel project revenue.
The kind of engineer who's been shipping since before most frameworks existed.
Bar widths are illustrative relative weights - exact allocations not public.
Backed by institutions on two continents - San Francisco's most competitive startup program and Oxford's oldest university fund.
Financial controllers who've lost weeks to pivot table maintenance. Auditors working across messy multi-entity Excel trails. R&D teams waiting for IT to respond to data requests. Procurement managers who know the answer is in the ERP somewhere but can't find the right module.
RISE, a financial consultancy, uses Zeit AI to analyze expenses across their entire organization - the kind of cross-entity work that would otherwise require custom scripts and several afternoons.
Zeit AI runs a forward-deployed model. Their engineers go to customers, work alongside business teams, and don't ship and forget. It's the Palantir playbook applied to an AI product: get in the room, understand the mess, make it work.
The company spans San Francisco and Gersfeld, Germany - small enough that everyone knows what everyone is working on, experienced enough that they don't need managers to tell them.
Leopold leads enterprise accounts across DACH. Marvin builds systems that generate €50M+ in revenue. Both accumulate 15+ years of combined data engineering frustration.
Leopold and Marvin leave Palantir and start Zeit AI, headquartered in San Francisco with a team also based in Germany.
Zeit AI is accepted into Y Combinator's record-breaking Summer 2024 cohort - the most AI-heavy batch in YC history, with 75% of companies building on AI.
Zeit AI presents at YC Demo Day, showcasing ZeitMind to investors and the broader YC network.
Y Combinator leads the round alongside Oxford Seed Fund, ACE Ventures, Stern Stewart Ventures, and others. Official launch published by YC.
Both founders worked at Palantir with a combined 15+ years of data engineering experience. They didn't study the enterprise data problem. They lived it, for years, inside the world's most demanding enterprise software company.
"Zeit" is German for "time" - a deliberate name for a product that saves enterprises the most expensive resource they have. Every week an analyst spends building reports manually is a week of insight that arrived too late.
Leopold von Waldthausen founded his first business, weview, at age 18 - and sold it before finishing his Computer Science degree at Oxford University.
Marvin Bornstein has been writing software for over 18 years professionally. He started coding before many of today's dominant frameworks and languages were mainstream - and he's shipped production code across chemicals, railways, media, utilities, and insurance.
Zeit AI is GDPR compliant and does not train its models on client data. In enterprise AI, that distinction matters enormously - it's the difference between a pilot and a blocked procurement.
The company operates across two continents - San Francisco and Gersfeld, Germany - reflecting the transatlantic nature of both founders' backgrounds and their target European enterprise customer base.