Nazlı Tan once asked a future colleague where she had learned set theory. There are easier ways to make conversation at work. Weather is popular. So is coffee. Tan reached for a branch of mathematics concerned with collections, membership and the foundations beneath the numbers. Beril, the colleague in that story, would become her chief of staff. The question offers a useful introduction to the co-founder and CEO of Hilbert: before getting comfortable with an answer, she wants to know what holds it up.
Her company applies that instinct to consumer growth. The subject sounds cheerful enough. More customers. More orders. A better month. Then come the qualifications: which customers, at what cost, buying again how often? Put those questions in different departments and a simple ambition acquires several competing definitions. Tan has spent much of her working life inside that complication. Hilbert is her attempt to build something underneath it.
A career spent following the calculation
Her education joins a bachelor’s degree in mathematics at Middle East Technical University, with a minor in logic, to a master’s in financial engineering at Boğaziçi University. The early jobs were correspondingly quantitative: credit analytics at Citi, followed by valuation, modelling and analytics at PwC. These are professions in which an attractive answer still needs a defensible method. A spreadsheet can be beautifully dressed and remain very wrong.
In 2018, she joined Getir as a growth manager. She became head of growth in 2020, then head of US growth in 2021. By 2023, her remit included US operations performance. Later came head of analytics at FreshDirect and VP of Growth at Flash Pack. The sequence moves from financial models to the behavior of people spending money, returning to a service, or quietly disappearing from it.
Read together, those roles suggest a career organized around the relationship between a calculation and a decision. The employers changed. So did the business settings. The recurring task was to understand a commercial system well enough to change something within it. That is a more demanding assignment than delivering a tidy monthly report. The report gets to stop at the end of the page.
Credit analytics
Growth
Analytics
Growth
Co-founder & CEO
Nine countries, one difficult job
At Getir, Tan built the growth function and scaled it across nine countries. That experience is central to the explanation of Hilbert offered by its investors. The geographic reach matters because a growth function has to operate inside the business it is meant to grow. A number that makes sense in one market needs scrutiny before it becomes someone else’s instruction.
There is an obvious temptation to treat expansion as the repeated application of a winning formula. The more interesting interpretation of Tan’s operating history is that it exposed her to the work required to make formulas useful. Data must be collected, definitions made consistent, and commercial questions translated into something that can actually be tested. International experience supplies more occasions for those assumptions to meet reality.
Hilbert’s own account of its origins describes scattered data, manually assembled answers and teams working from different versions of events. It also describes decisions arriving late. These are unglamorous problems with an unfortunate talent for surviving expensive software purchases. A company can buy another dashboard much faster than it can settle an argument about what belongs on it.
“A system built by operators, mathematicians, and engineers who lived the complexity of growth”
Hilbert’s description of its founding approach
Four colleagues build the missing layer
Tan co-founded Hilbert with Ceyda Erten, Özgür Akaoğlu and Cenk Batman. Their experience combines growth execution, operations, machine learning and data engineering. The founding group had worked on growth systems at Getir and FreshDirect. ScaleX Ventures backed the company at the pre-seed stage and returned as an investor when Hilbert raised its Series A.
That mix of disciplines is part of the company’s argument. A useful growth system needs the machinery that handles information and the judgment to understand what an operator will do with it. Tan’s place in the group comes from the growth side of that relationship. Erten brings operational experience, alongside Akaoğlu’s AI and machine-learning focus and Batman’s engineering role.
The name supplies another clue to the ambition. It refers to David Hilbert and the effort to establish secure mathematical foundations. Naming a company after a mathematician is a fairly economical way to warn everyone that definitions may be discussed. Here the connection is practical: the product’s starting point is a shared structure for interpreting the business before asking agents to make decisions within it.
The agent has boundaries
Hilbert presents its software as a shared model of a consumer business, connecting information across functions and identifying actions that might improve growth. Its public product explanation starts with data, then moves through analysis toward execution. The order is consequential. If two departments mean different things by a returning customer, asking an agent to move faster gives the disagreement a motor.
In Tan’s explanation, deep-learning models establish predictions from customer history, while language-model agents operate within those boundaries. The agents are restricted from doing the mathematics beyond rounding. That is a design claim about how Hilbert allocates work between models. It also reveals something about the founder’s priorities: a fluent answer needs a reliable calculation beneath it when a budget depends on the result.
The distinction can sound fussy until money enters the picture. A recommendation about customer value can influence which audience gets an offer, how much is spent attracting it and which existing customers receive attention. Accuracy has consequences beyond the screen. The mathematical foundation is therefore part of the commercial proposition, alongside the ease of asking a question in ordinary language.
The grocery aisle has its own rules
Tan also questions the assumption that Silicon Valley’s consumer growth playbooks travel comfortably into every consumer enterprise. She defends paid acquisition as a legitimate route to profitable customers. A retailer cannot necessarily replace advertising volume with organic discovery simply because that sounds more elegant. Her point concerns the economics of the customer over time, rather than the prestige of the channel that first introduced them.
The businesses associated with Hilbert make that distinction concrete. Its public customer roster includes Walmart, FreshDirect, Blank Street and Levain Bakery. Groceries, coffee and cookies involve repeat purchasing in the physical world. The businesses share the problem of understanding customer behavior, while operating in categories with different purchasing rhythms. A common analytical foundation still needs to accommodate those differences.
For Tan, this is familiar territory. Getir and FreshDirect put consumer demand close to operational reality. Her current work returns to that intersection from the position of a software founder. The interpretation suggested by her career is straightforward: a growth decision becomes more useful when it can be understood in relation to the rest of the business. A campaign does not exist in a separate universe, however convenient that would make the meeting.
Now the system has to travel
Hilbert announced a $28 million Series A led by Andreessen Horowitz in April 2026. By September, Dana Grabiner, previously Just Eat’s head of grocery, had joined as head of go-to-market to lead the UK launch. The appointment extends the company’s connection to operators with experience in consumer commerce. It also moves Tan’s work into another phase: bringing the system to more businesses.
Her professional circle now includes the a16z Growth Engineer Fellowship, which listed her among its founding fellows and industry leaders in 2026. Hilbert’s hiring has included an AI Growth Architect tasked with testing the product on real growth problems, challenging its reasoning and feeding improvements back into the system. The job description makes scrutiny part of the work itself. Someone is being hired to ask difficult questions of the machine.
There is a public teaching thread, too. Hilbert Space has featured Alex Schultz in a conversation about growth foundations, measurement and experimental discipline. The series brings those questions into discussions with operators rather than leaving them inside a product demonstration. For a company built around agreement on what numbers mean, that is a fitting subject. It gives the surrounding work a vocabulary: active users, retention, bias and the quality of an experiment. Those are also questions a growth team can carry back to its own business, whether or not it buys new software.
Tan’s story makes that role feel appropriate. It began, for one colleague, with an unexpected question about set theory. It now concerns how companies define customers, evaluate behavior and decide where money should go. The continuity is the insistence on examining the foundation before becoming satisfied with the answer. A green chart is a pleasant thing to bring to a meeting. Tan has built a company around the questions that ought to follow.