Quantitative & AI-based analytics, low-latency infrastructure and algorithmic automation for data-driven domains
My research centers on applying quantitative methods to massive, high-frequency datasets. This work extends into automated decision making, autonomous agents and algorithmic automation & AI as well as the low-latency infrastructures required to execute decisions in real time. I focus on the intersection of applied research and industry-academia cooperation, with particular interest in transfer and venture creation.
Selected publications & working papers
Peer-reviewed research papers and academic contributions exploring data analytics, algorithmic automation, and autonomous decision systems.
Hermes: An Autonomous Algorithmic Trading Agent for Cross-Exchange Spot Markets
Capturing price discrepancies across different venues requires continuous market monitoring of massive data, automated decision making and low-latency execution. In this work we present Hermes, an autonomous algorithmic trading agent built on a low-latency execution infrastructure. It integrates a rule-based decision cycle, low-latency order execution and automated asset rebalancing within a continuous feedback loop. Thereby, Hermes evaluates and sizes trades based on depth, fees, profit targets, and constraints, automatically rebalancing assets and reassessing opportunities before further execution, enabling repeated exploitation of opportunities with significant volume without human intervention. The agent was deployed in a real-world professional environment, executing more than 400,000 transactions.
QAVCM: A Quantitative Aggregation and Analytics Platform for the Fragmented Voluntary Carbon Market
QAVCM addresses the structural fragmentation of the voluntary carbon market. The integrated analytics and execution platform harmonises certificate data drawn from over 1.2 billion data points spanning 11,599 projects, representing 1,093,056,213 tCO₂e of tradable credits and 2,944,950,676 t/year of reduction capacity. Machine learning estimates fair values and explainable AI using SHAP makes model outputs inspectable. A smart order routing layer compares connected venues and splits orders to minimise execution costs.
Algorithmic Entrepreneurship: Evidence on Data-Driven Algorithmic Automation and Output-to-Employee Scaling in a Digital Asset Market Infrastructure Venture
This study examines how Data-Driven Algorithmic Automation shapes Output-to-Employee Ratios and scalability in a Digital Asset Market Infrastructure Venture. It benchmarks publicly disclosed and blockchain-based data from large-scale algorithmic infrastructures (Tether, Sky Protocol, Hyperliquid) against an empirical analysis of a SaaS venture's client transactions, volumes, fees and staffing levels from 2021 to 2024, finding that Output-to-Employee ratios increased disproportionately over time. By emphasizing operational throughput over revenue-based metrics, the study contributes to debates in Automation Economics, Data-Driven Business Models and Labor Productivity.
Executing Arbitrage at Scale: Empirical Evidence on Crypto Spot Market Efficiency
An assessment of proprietary live execution data generated by a commercial low-latency algorithmic decision and execution engine. In 2021, the underlying low-latency infrastructure was licensed to professional quantitative market participants and processed the algorithmic execution of more than 430,000 spot transactions across eight exchanges and 32 markets, representing a spot volume of more than USD 450 million. Our findings provide evidence from real-world transaction data on how crypto spot arbitrage opportunities materialize, vanish, and can be executed, ultimately shaping efficiency in spot markets.
DePIN for Enhancing the Resilience of EWF Infrastructures: Systematization and Potential for the Federal State of Brandenburg
Critical infrastructures in the areas of energy, water and food (EWF) require new resilience concepts in the face of increasing global challenges. This article examines how decentralised physical infrastructure networks (DePIN) can contribute to increasing the resilience of EWF systems by combining blockchain, the Internet of Things and artificial intelligence, focusing on the Brandenburg region. The results show that DePIN approaches reduce single points of failure and enable a faster, autonomous response to disruptions, thereby sustainably improving regional security of supply.
Towards Infrastructure Digital Twins as Innovation Enablers in Rural Energy Systems: Early Developments for Renewable Transitions in Brandenburg
Rural energy systems are central to renewable energy transitions but face persistent innovation challenges related to spatial dispersion, fragmented infrastructure ownership, and limited coordination capacity among actors. While rural regions often host substantial renewable generation potential, early-stage infrastructure innovation is constrained by heterogeneous data sources and a lack of shared situational awareness. This Research-in-Progress paper presents early developments from the Resilient Infrastructure Technology Suite (RITS) project in Brandenburg, examining how Infrastructure Digital Twins (IDTs) may function as innovation enablers in rural energy systems. The IDT is developed as a modular, GIS-based analytical and visualisation prototype supported by AI-assisted exploratory analytics.
Athena: Smart Order Routing on Centralized Crypto Exchanges Using a Unified Order Book
In liquid markets based on order books, the price achieved for buying and selling deviates only slightly from the assumed reference price, meaning trading is associated with low implicit costs. This paper describes the design and implementation of Athena, a smart order routing algorithmic system that automatically splits orders across multiple exchanges to minimize implicit costs, collecting and merging order books from several centralized crypto exchanges into an internal unified order book.
Short Paper: A Low-Latency System for Collecting Massive Crypto Option Tick Data from Deribit
As digital asset markets mature, options are gaining increasing importance. Deribit has emerged as the dominant exchange for crypto options, accounting for the majority of trading volume. To facilitate backtesting and market analysis, access to high-quality historical data is essential. This paper introduces a low-latency system for collecting massive volumes of tick data together with the best bid and best ask from Deribit, capturing every price change with precise timestamps to provide granular insights into market microstructure and trading dynamics.
Is Real Estate Tokenization the Killer Application?
This chapter shows how tokenization digitally depicts real assets such as real estate and makes them flexibly tradable. Security tokens in particular offer investors security and could establish themselves as a standard in the real estate industry. Tokenization makes real estate more mobile, liquid and accessible, lowers entry barriers for small investors and enables automated transactions with smart contracts.
A Low-Volatility Strategy Based on Hedging a Quanto Perpetual Swap on BitMEX
In 2016, BitMEX introduced a novel type of derivative — Perpetual Swaps, futures with an infinite term. Perpetual swaps provide a new strategic risk management tool for digital assets due to their custody-free nature, high leverage, and funding mechanism, but there has been little quantitative analysis of their benefits. This paper introduces a systematic strategy that combines a Quanto Perpetual Swap with a spot position to benefit from the funding mechanism.
Hephaistos: A Management System for Massive Order Book Data from Multiple Centralized Crypto Exchanges with an Internal Unified Order Book
This paper introduces Hephaistos, a system designed to collect, process, and unify order book data from 22 centralized exchanges and 55 currency pairs. Unlike tick data, which only shows the last transaction, order books provide a more accurate view of market prices and available volume. Hephaistos aggregates this data into a Unified Order Book, enabling a Smart Order Routing algorithm that splits orders across multiple exchanges to achieve better execution prices. The system captures around 32% of global daily spot trading volume, and tests show that the algorithm reduces slippage, making it effective for high-frequency trading.
Real Estate Tokenization in Germany: Market Analysis and Concept of a Regulatory and Technical Solution
Real estate is the largest asset class and is popular with both professional and retail investors. However, it suffers from illiquidity and high equity entry barriers. The 2021 adoption of the Electronic Securities Act by the German Bundestag established a legal framework for tokenizing real estate assets and their management via digital ledger technology in Germany. This paper outlines a business concept for managing ownership and transactions in German real estate using blockchain, and includes a market analysis of existing approaches and legal limitations specific to Germany.
Examining Liquidity of Exchanges and Assets and the Impact of External Events in Centralized Crypto Markets: A 2022 Study
Most cryptocurrencies are traded on centralized exchanges using order books. Liquidity, alongside trading fees, is a key factor in exchange choice. This study analyzes liquidity across nine order books (three pairs on three exchanges) over 2022, showing that higher market cap and trading volume lead to greater liquidity. Liquidity correlations between exchanges are moderate. External events like the FTX and Terra Luna crashes significantly impacted liquidity, with recovery time depending on exchange size and market cap.
OrderBookVis: A Visualization Approach for Comparing Order Books from Centralized Crypto Exchanges
Trading for a currency pair on centralized exchanges is organized via an order book that collects all open buy and sell orders and serves as the basis for price formation. While basic 2D visualizations are typically provided, they do not support comparing different order book snapshots. This work introduces OrderBookVis, a 2.5D visualization that displays multiple order book snapshots in sequence on a 2D reference plane. Use cases include analyzing the temporal evolution of an order book for a specific market and comparing different order books across multiple markets.
Research was conducted in association with University of Potsdam Hasso Plattner Institute and XU Exponential University.
Request Full Citations & Code