Consultant for Citibank, Senior Data Scientist – 2020-Present
- Created an antivirus server that formed the backbone of the file security for the Chief Data Office. Pushed this to SIT, UAT, Prod and COB. It was the first production release for the CDO for 2021.
- Wrote a template system for supplementary uploads into the MongoDB for the CDO, expediting this process on the front and backend by over 10x.
- Implemented Neo4j graph databases to encompass firewall rules which contributed to an overall model of security infrastructure.
- Optimized server configuration and data insertion scripts in Cypher to reduce load time into our Neo4j database from one day to less than one hour.
- Developed novel algorithms for reconciliation of firewall rules across disparate systems that reduced time latency from up to two days to less than one hour.
- Developed novel algorithms to detect loops within functional graphs which facilitated the development of our Data Lineage project.
Ghost Squawk, CTO and Co-Founder 2018-Prese
- Released a deflationary token (source code availantble on GitHub) using blockchain development tools including truffle, ganache-cli, remix and more.
- Preparing to release an ERC-20 token. Verified the security and robustness of this token by writing test cases leveraging OpenZeppelin, and deploying on the Ropsten and Rinkeby test networks.
- Implementing the ERC-721 standard to release Non Fungible Tokens for art and legal contracts.
- Developed Natural Language Processing implementations using scikit-learn and python-nltk to determine if content is crypto-related with a success rate of over 99%, which formed the backbone of the news platform.
- Wrote a web-scraping service using python with selenium and docker to monitor major financial regulatory entities from around the world to identify when they released crypto-related legislature.
- Created proprietary technical indicators in Nodejs to identify liquidations and large block order activity, which were implemented in trading algorithms that earned a monthly profit of up to 10%.
Institute of Trading and Finance, Instructor and CTO 2014-2018
- Developed the curriculum for Technical Trading and Quantitative Finance, and taught students how to successfully trade various asset classes, including commodities, stocks, bonds, currencies and more.
- Spearheaded initiatives to modernize infrastructure and coordinated the company website development (www.iotaf.com), which resulted in the creation of a successful online program, and boosted revenue by 15%.
Go To Technologies, Consultant and CDO 2015-2018
- Implemented technical and fundamental strategies using R and Python on Azure to trade various securities and achieve a 0.5% average daily return.
- Wrote a portfolio management algorithm that has generated up to 3% quarterly returns on U.S. stocks.
- Used Natural Language Processing and web-scraping to intelligently match candidates with open positions, which improved the efficiency of our placement process by 70%.
Palo Alto Networks, Senior Data Scientist 2014-2015
- Unveiled state of the art 3D graph rendering technologies using WebGL and ThreeJS at Black Hat 2015, which enabled us to visualize thousands of malware relationships.
- Improved the company’s major product Wildfire by leveraging machine learning algorithms written in R and Python to identify false positives, reducing the rate of false positives by 90%.
- Implemented Hadoop, HBase, and Pig to enable large scale processing of PDNS data.
- Applied NoSQL database and graph database technologies like MongDB, Neo4j, and Aurelius Titan DB to analyze billions of potential malware relationships.
Kovach Technologies, Consultant 2013-2014
- Essential contributor for an in-flight entertainment product written in C++ on a CentOS OS.
- Designed embedded applications using IBM Rhapsody and GNU tools on Linux.
- Hardened applications using emulation technologies to simulate real life flight scenarios.
- Performed security audits to improve overall robustness.
Raytheon/SI, Machine Learning Specialist 2008-2013
- Spearheaded numerous innovative projects by leveraging a strong background in Machine Learning and Artificial Intelligence.
- Completed contract-based and R&D projects leveraging a mastery of C and C++, with a focus on Linux.
- Implemented system engineering plans, interfaced with customers, and contributed to overall project management to bring projects success and ensure customer satisfaction.
- Formed the backbone of Raytheon’s corporate firewall, Rshield by writing a custom high-performance Neural Network library.
NASA FSGC Grant, Analyst 2007-2008
- Invented a novel, effective signal processing algorithm to identify characteristically harmful vibrations which improved the overall quality of the mission.
- Publicly released this technology and its applications to financial analysis at the World Conference of Nonlinear Analysts on July 7th, 2008.
Florida Institute of Technology 2003-2008
- MS Applied Mathematics
- BS Mathematical Sciences
- BS Physics
- Graduated Magna Cum Laude
Skills and Research Interests
- Proven track record of mastery in the following languages: Python, R, C, C++, Mathematica, Matlab, x86 Assembly, and FORTRAN.
- Publications and research specialties include Artificial Intelligence, Machine Learning, Parallel Computation, Mathematical Modeling, Graph Theory, and Visualization.
- Wrote a high frequency trading engine in Linux using a Neural Network library written in x64 assembly which leveraged the SSE instruction set to outperform competing systems by up to an order of magnitude.
- Utilized tools such as Gnuplot, R, Python, ThreeJS and WebGL to create visualization frameworks for large and complex data.
Honors and Publications
Notable publications (see https://www.researchgate.net/profile/Daniel_Kovach2 for a full listing):
- Kovach, Daniel J. “The Computational Theory of Intelligence: Information Entropy.” International Journal of Modern Nonlinear Theory and Application, 3, 182-190.
- Kovach, Danny J. “The Functional Analysis of Non-Archimedian Numbers.” Nonlinear Analysis and Phenomena 3 (2006): 35-56.
- Lakshmikantham, Vangipuram, and Danny J. Kovach. “The Hybrid Real Number System.” Hybrid Systems and Applications 1 (2007).
- Tenali, Gnana B., Danny J. Kovach, and Vangipuram Lakshmikantham. “The Hybrid Set Theory.” Hybrid Systems and Applications 1 (2007).
- “Information Theory in Quantitative Finance”, BJMT Alkalmazott Matematikai Konferencia, June 2nd, 2016
- “The Computational Theory of Intelligence”, ICNAAM 2013, September 26th, 2013
- “Yet Another Heap Spray Detector”, Schmoocon, January 29th, 2010
- “Artificial Intelligence Concepts”, Raytheon/SI Lecture, on October 28th, 2008
- “Vibration Control Analysis, A Signal Processing Approach”, World Conference of Nonlinear Analysts, on July 7th, 2008
- Open Source software contributions include Mathematical and Machine Learning libraries written in C and C++, with Python bindings available for download at https://github.com/quant-guy.
- Received an award for outstanding performance at Raytheon, in June of 2010.