At its core, Runtime reflects the backgrounds of Deniz Erkan and Raed Malhas: two technology builders who first met at Microsoft, spent the early part of their careers around mission-critical infrastructure, built operating companies together, and later turned that same systems mindset toward investing.

Technical Foundation

Deniz holds Computer Science and MBA degrees from METU in Turkey, the MIT of Turkey of the 90s. Admission required ranking near the very top of a central nationwide exam; Deniz ranked in the top 250 out of roughly 2 million people.

In 1999, Microsoft visited the METU campus in Ankara and selected 2 people from a graduating class of 50. Deniz was one of them. He moved to Seattle and worked on SQL Server Enterprise, Microsoft’s enterprise database platform built to compete with Oracle and IBM DB2. SQL Server Enterprise was Microsoft’s second-largest codebase after Windows Server and powered mission-critical systems for banks, airlines, insurance companies, and other institutions where reliability was not optional.

That work shaped Deniz’s engineering foundation: infrastructure software with zero tolerance for failure, 99.99% uptime expectations, and production environments where a crash could affect real businesses at scale.

Deniz met Raed at Microsoft. While Deniz worked on SQL Server Enterprise, Raed was building Microsoft’s global payments platform: a high-volume, compliance-heavy, reliability-critical system that started as a small team effort and later became one of Microsoft’s core transaction platforms. Raed’s own background was also rooted in systems engineering: he earned his degree in Computer Engineering from Purdue University and today serves on the board of Purdue’s Computer Engineering department.

That early Microsoft experience matters because Runtime is also infrastructure. It is not just a model, a backtest, or a discretionary trading process. It is a live system that has to ingest data, understand positions, evaluate risk, calculate option structures, place orders, and keep running under real market conditions.

Entrepreneurial Experience

After Microsoft, Deniz and Raed built technology companies together. One of those companies reached almost $50 million in annual recurring revenue at its peak.

The more important point is that they built a profitable, cash-flowing technology company. That is a different discipline from simply chasing top-line startup growth. It requires financial control, unit-economic discipline, operational focus, and the ability to make hard tradeoffs over long periods of time.

Their investors got their money back with strong returns. Several of them later committed to Runtime Fund as LPs, not because they understood every detail of the strategy on day one, but because they had known Deniz and Raed for years, had seen them operate under pressure, and trusted their judgment as builders.

That company-building history is part of Runtime’s DNA. It was built by operators who had already taken complex systems from idea to production, revenue, profitability, and investor returns.

Personal Investing Experience

Deniz began investing seriously in 2003 while living in Seattle and working at Microsoft. He opened a Fidelity account and approached investing with an old-school philosophy: stay inside the circle of competence, identify strong companies, own the winners, and let them compound.

He read Warren Buffett’s annual letters, Charlie Munger, and the broader value-investing canon. More importantly, he applied those principles in his own account over a long period of time.

Over 23 years, Deniz’s personal account compounded at 19.86% annually, producing a total return of 6,767%. That personal investing record is an important part of Runtime’s foundation because Runtime was built from lessons learned through decades of real capital allocation, not from a theoretical research exercise.

The first lesson was that old-school investing works. The second lesson was that it leaves important problems unsolved: position sizing, when to add, when to trim, how to handle drawdowns, and how to keep emotions from overriding a good thesis.

Two Schools of Thought

In 2009, Deniz had a major investing insight after reading Warren Buffett’s discussion of selling long-dated index put options during the 2008 financial crisis. Until then, he associated options mainly with leverage and danger.

That changed his thinking. If Buffett could use options as a risk-management and return-shaping tool, then options were not just speculative leverage. They could be instruments for structuring asymmetric outcomes.

That realization led Deniz into a roughly 10-year study of options, volatility, Greeks, term structure, payoff asymmetry, and portfolio construction. During that period, he continued investing with the same old-school philosophy while experimenting with covered calls, straddles, butterflies, and other option structures around stocks he already understood.

In 2019, a second insight came from Jim Simons and Renaissance Technologies. After reading “The Man Who Solved the Market,” Deniz reconsidered his view of algorithmic investing. He had previously believed algorithmic strategies could work for a few years and then stop working. Simons showed that data, models, risk control, and position sizing could compound for decades if the system was built correctly.

The timing mattered. In 2019, Deniz and Raed had also built an AI phone product focused on eradicating spam calls and protecting people from scams. The product labeled suspicious calls and was designed to answer calls and speak like a human. This was before ChatGPT became part of everyday life, so the team had to build much of the infrastructure from first principles: neural nets, speech synthesis, real-time inference, big data pipelines, and high-volume signal processing.

The system handled close to a billion phone numbers and more than 100 million incoming real-time phone-call signals per day. That gave Deniz and Raed practical infrastructure in big data, AI, and neural networks before those tools became mainstream.

By then, the pieces were in place: old-school investing, options expertise, AI infrastructure, and a systems-builder mindset.

The Runtime Thesis

Runtime came from a simple question: what if you keep the best part of old-school investing, the ability to identify strong companies and let them compound, but systematize the parts where human judgment often breaks down?

Picking the winners is only part of the problem. The harder questions are how large the position should be, when to buy, when to sell, how to react to volatility, and how to preserve upside while limiting the damage from mistakes.

Runtime’s core idea was to combine fundamental stock selection with an AI-powered options overlay. The stock portfolio provides the old-school foundation. The options layer is designed to reshape the return profile: boost winners asymmetrically, limit downside on losers, and make position sizing and risk decisions more systematic.

This is the origin of Runtime’s old-school plus new-school identity. The old school is the investing foundation: circle of competence, company quality, compounding, and patience. The new school is the AI-powered infrastructure layer: data, options, automation, real-time calculation, and systematic risk control.

Runtime Goes Live in Deniz’s Personal Account

In 2019, Raed encouraged Deniz to build the platform, telling him to focus on it while Raed handled the operating company. Deniz stopped working day to day on the company and repurposed the AI infrastructure they had built into what became Runtime’s investing platform.

It took about a year to build the first live version.

In April 2020, Deniz launched Runtime v0.1 inside his Interactive Brokers account. The platform integrated with IB’s APIs.

Deniz viewed the platform as infrastructure software in the spirit of SQL Server. The engine had to work. It could not crash at the wrong time. A failure was not just lost uptime; it could mean lost money, distorted strategy behavior, or incorrect feedback into the next round of system improvements.

The early phase was about stabilization. Live production exposed the kinds of problems that do not appear in a static model: streaming interruptions, freezes, data issues, memory pressure, crashes under load, and edge cases that only appear when real markets are moving.

Deniz initially funded Runtime with about $110,000 of his own money. As the system stabilized and his confidence increased, he added capital in stages every few months. Raed did not invest at that stage, and neither did outside investors, because the platform was still being proven.

Over the next 5.5 years, Runtime compounded at 80% annually in Deniz’s personal account, producing a total return of roughly 2,440%. This was live capital in a personal account, not fund performance, and it became the proving ground for the system before Runtime was launched commercially.

From Personal System to Commercial Fund

After the first live version launched in April 2020, Runtime continued to evolve through years of real market exposure. The platform was tested against live prices, imperfect data, execution issues, changing liquidity, operational incidents, and actual capital at risk.

That production experience matters. Runtime’s commercial platform is not simply the original personal-account system repackaged for outside investors. It is a substantially more advanced implementation of the same architecture, with institutional-grade infrastructure, live execution systems, position reconciliation, monitoring, and disciplined risk controls built into the strategy. The commercial version adds product-specific calibrations and stricter controls around position sizing, liquidity, volatility, drawdowns, and exposure.

The scale of the infrastructure is part of the moat. Options are not just a trade idea; they are a data and systems problem. Historical option chains can grow from tens or hundreds of terabytes into petabyte-scale infrastructure once higher-resolution data, indexing, derived features, simulations, backups, execution records, reconciliation, and monitoring are included. An outside observer may understand the broad strategy, and AI can help improve parts of the system, but intelligence alone is not enough. Without the private data, live execution history, portfolio context, and production feedback loop, there is little for even a very capable model to compound on. Deniz expands on this idea in Why Options Infrastructure Is Hard to Copy.

That iteration process became one of Runtime’s core architectural principles: a self-improvement inner loop. Runtime was not designed as a static model researched once and left alone. It was built as a live system where production behavior creates feedback: market data, execution quality, portfolio outcomes, risk signals, operational incidents, and edge cases all become inputs into the next version of the platform.

This is the same idea Deniz later described in The New Moat Is the Inner Loop: private context, disciplined action, measured outcome, and improved future behavior. Runtime has been compounding through that loop since its first live deployment.

That evolution is central to the story. Runtime was not built by backtesting an idea and immediately raising a fund. It was built as live infrastructure first, refined through years of production use, and only then turned into a separate commercial fund.

Deniz and Raed launched Runtime Fund on April 1, 2026, with NAV as the fund administrator and approximately $20.2 million in AUM, anchored by approximately $18.5 million of Deniz’s own capital. Since launch, as of June 20, 2026, AUM has grown to approximately $30 million, with a small group of outside LPs, primarily people who knew Deniz and Raed from their prior entrepreneurial and company-building career, joining with modest initial commitments. From the beginning, Runtime has been built with founder alignment at the center: the strategy is not abstract research capital; it is capital Deniz lives with personally.

As Runtime moves from personal system to commercial platform, the first product architecture is being organized around two investor profiles.

Runtime Core is designed for family offices, qualified high-net-worth individuals, and long-term private investors who understand the possibility of larger interim drawdowns and want fuller participation in the platform’s return-seeking engine.

Runtime Low Volatility is designed for institutional capital allocators who want the same underlying research, execution, and risk infrastructure in a more conservative calibration, with greater emphasis on volatility control, drawdown management, and smoother participation through market cycles.

The important point is that these are not two separate systems. They are two calibrations of the same architecture: the same research engine, the same live trading infrastructure, the same risk system, and the same self-improvement loop applied to different investor objectives.

The longer-term vision behind this architecture is described in Runtime’s manifesto, What We Stand For. Runtime exists to build a new investment architecture for a world where human judgment, machine intelligence, computation, and capital are becoming more deeply intertwined.

The fund is the commercial expression of the same system-building path: mission-critical engineering from Microsoft, profitable company-building, long-term personal investing, options expertise, AI infrastructure, and years of live trading experience brought together into one investing platform.

Skin in the Game

Runtime was built around a simple principle: alignment should be real, with approximately 90% of Deniz’s net worth invested alongside LPs rather than building a firm whose main economic engine is fees and AUM. He wrote more about that owner-operator alignment in Skin in the Game.

Deniz has lived in New York since 2007, where he and his wife are raising their two children.

Continue Reading

For readers who want to go deeper, these essays expand on the main ideas behind Runtime:

• Skin in the Game — why founder alignment and personal capital are central to Runtime.
• Why Options Infrastructure Is Hard to Copy — why options investing depends on serious data, compute, and execution infrastructure.
• The New Moat Is the Inner Loop — how private context, action, measurement, and improvement become a compounding system.
• What We Stand For — Runtime’s broader investment philosophy and long-term architecture.