PETER ZAK
Personal Portfolio
About Me
I am a finance-oriented business student passionate about data science and predictive modeling for algorithmic trading. I am particularly interested in building automated data science solutions to facilitate the process of informed and quantitative oriented decision making, both in trading and corporate governance.
Education: ESCP Business School
Bachelor of Science in Management (Finance Major)
(Graduation: April 2025)
The world’s oldest Business School (est. 1819)
“Grandes écoles are elite academic institutions that admit students through an extremely competitive process, and a significant proportion of their graduates occupy the highest levels of French society. Similar to Ivy League universities in the United States or Oxbridge in the UK or the C9 League in China.”
ESCP’s Executive MBA ranks 2nd worldwide in Financial Times 2024 ranking
.
4th Best Business School in Europe (Financial Times)
.
One of the “trois Parisiennes“
It is known as one of the trois Parisiennes
(three Parisians), together with HEC Paris and ESSEC , designating the three most prestigious business schools in France.
ESCP Website: https://escp.eu
Financial Times Ranking (1st master in finance worldwide): https://rankings.ft.com/business-education/masters-in-finance
Financial Times Profile: https://rankings.ft.com/schools/115/escp-business-school/programme-portfolio
Programming Languages & Technical Tools
Python
– Libraries: Pandas, NumPy, Matplotlib, TensorFlow, SciPy, Keras, PyTorch, Seaborn, scikit-learn, Plotly, TA-Lib, etc.
– Data analysis, statistical modeling, machine learning, and automation
– Experience with API integration, data visualization, and workflow optimization
– Projects: algorithmic trading programs, micro market regime detection models (ML), object detection, web scraping, automation solutions
C / C++
– Learned mainly to focus on low-latency execution, concurrency, and memory optimization
– Used for HFT solutions and performance-critical applications
– Deployed a HFT trading program exploiting price differences between BTC/USDT and BTC/USDC (statistical arbitrage)
Experience with multi-threading, real-time data processing, and system-level programming
SQL
– Learned SQL language and commands to extract, transform, load (ETL) large datasets
– Learned queries, joins, subqueries, data aggregation
– Database design, optimization, and performance tuning
Linux Commands & Shell Scripting
– Used to run programs, manipulate environments, automate workflows
– shell scripting, CLI commands, task scheduling (cron)
– System monitoring, process management, performance optimization
– Used for deploying trading algorithms and automating data pipelines
HTML
– Analyzed website structures for web scraping
– DOM manipulation, data parsing, and extraction techniques
– Integrated with Python tools for automation and data collection
– Applied in market research and competitive analysis
Other Technical & Visualization Tools
Microsoft Power BI
Microsoft Excel
(Advanced Level)
Tableau
VBA
PROFESSIONAL EXPERIENCE
CISCO’s LIFT
“Leaders in Finance & Technology program is a 10-week finance summer internship experience aimed at identifying and developing future finance leaders and CFOs”. Only a total number of 8 Interns were hand-picked for this internship from the whole EMEA region. During the Internship I:
Link to CISCO’s website: https://www.cisco.com
My CISCO experience:
– independently automated analytical workflow by creating interactive dashboards using Power BI and Power Query to streamline the process of identifying and filtering financial data. Whenever a new raw dataset is uploaded, my dashboards automatically update themselves by applying predefined data transformation steps and allow users to retrieve information of interest by applying customized easy to use filters.
– collaborated in a team of 3 other LIFT Interns by analyzing the correlation between discount and revenue levels of best selling EMEA products. Our team provided recommendations and findings before CISCO’s EMEA executives panel and ~200 Cisco employees, followed by a Q&A session.
– worked on Cisco’s platforms & tools, exporting datasets, analyzing trends over years and across product families, drawing conclusions and suggesting improvements.
– gained practical experience and proficiency in using tools like Excel, Power BI and Power Query for financial data analysis (financial records from previous accounting periods, gross-to-net-revenue walk).
– applied quantative modeling techniques and concepts learned during my extracurricular online specialization courses at Wharton University (Finance & Quantitative Modeling for Analysts Specialization”) to real corporate financial data.
Algorithmic Trading Research Project (FlashTrader)
I built a complete prototype – an algorithmic trading program – that executes transactions and rebalances the portfolio every minute for optimal performance based on ML generated price predictions.
Then I created a blueprint for a potential start-up: “FlashTrader”, which was validated by my university as an 8-week professional experience. I created an entire business plan: marketing strategy, business model canvas, legal considerations, market segmentation, cost analysis, etc.
More information on the Quantitative Trading Model:
– Trained ML models using OHLC prices, volume patterns and numerous technical indicators, including: MACD, RSI, MFI, Bollinger bands, ATR and EMA;
– Used Binance API to trade cryptocurrencies trading pairs and their derivatives using the platform, then deployed my model on AWS EC2 (Elastic Compute Cloud) to ensure reliable continuous 24/7 background operation in the cloud.
-Used libraries like: TensorFlow, Pandas, Pandas-ta, Numpy, Matplotlib, Scikit-learn, Seaborn, Joblib, Pytz, etc.
-Trained and tested multiple models using Random Forest / Gradient Boosting Machines (XGBoost) / LSTM / feature based, employed the “adam” optimizer, experimented with various estimators, random_state, max_depth, saved my models using “Python Pickle” and used techniques to avoid overfitting, such as dropout layers, L2 (Ridge) regularization and k-fold cross-validation. I later embedded these models to predict prices into a trading program, which specifies the trading logic
-Implemented overall trading logic: portfolio rebalances on a minute-to-minute timeframe based on the predicted prices of each potential asset. The weight assigned to each cryptocurrency is equivalent to that assets fraction of total differences in price for all assets, so that assets with bigger predicted price movements are assigned higher weight.
Back-tested my model:
• Average annual return: 33.7%
• Sharpe Ratio: ~2.1
• Annual Standard Deviation: ~0.16
• Maximum Drawdown: 7.3%
-Manipulated and cleaned datasets: replaced missing values with averages, removed NaN values, outliers using z-scores, etc.
– Created special technical indicators to deploy different strategies. For example, I created an EMA based, double Bollinger bands for micro trend regime switching strategy. During a bullish trend, I buy when the price tags +1 SD above EMA and sell when it tags +2 SD. I also set the Take Profit (TP) at +2SD and Stop Loss (SL) at EMA level. Furthermore, I can short when it tags +2SD and set the TP price at +1 SD. During a bearing trends, I apply the opposite logic and short when the price tags -1SD and buy at -2SD. In case of a sideways market, I just long at -1SD and short at +1SD.
Using EMA (Exponential Moving Average) instead of SMA (Simple Moving Average) makes the indicator more responsive to recent price changes and more appraise for high-speed algorithmic trading, for example on a minute to minute intervals.
JET ESCP
JET ESCP is the oldest consulting Junior Enterprise established at ESCP in 2004 that serves as a non-profit civil social organization formed and manage exclusively by students. It provides services to companies, NGOs and other institutions.
Link to JET ESCP’s website: https://www.jetescp.com
My JET ESCP experience:
– Participated in numerous projects, including an urban transportation sustainability initiative, startup accelerator, inclusive education project, etc.
– Analyzed data, conducted market research, offered solutions to clients to streamline inefficiencies and generate additional revenue streams;
– Acted as a project leader for an environmental sustainability development initiative, managed a group of 4 consultants, held weekly meetings, assigned workload, presented findings in front of clients and supervisors.
ESCP STUDENT AMBASSADOR
I actively represented ESCP Business School, not only during open days or information session, but proactively helped organize seminars, responded to countless e-mails by giving advice and counsel to potential candidates and their parents.
Bachelor’s Thesis
Thesis Title: “Exploring Alternative Data Sources for Quantitative Trading in the Technology Industry”
Research Questions:
1) Which publicly accessible alternative data sources have the potential to predict future stock prices or company earnings in the technology industry, and what is their predictive capability?
2) How to incorporate them into a systematic investing analysis process from the quantitative standpoint?
3) How can retail investors access and make use of alternative data to improve their investment choices and make more informed investment decisions?
Alternative Data Used
– lithium carbonate, cobalt, REE, etc. prices to predict the profit margins of companies including EV manufacturers
– Amazon product reviews analyzed by sentiment
– Satellite Imagery of parking lot of manufacturing facilities analyzed with CNN to predict the output for a fiscal quarter
– social media & news feed sentiment analysis for algorithmic trading
– app download statistics to predict the companies are gaining market share quicker than others considering the market growth rate as a whole
Methodology
Quantitative approach; focus on comparing ML models performance, like LSTM and tree-based models ,with and without inputting various alternative data sources to predict the target variables
LANGUAGES
English
Level: Native
Polish
Level: Native
German
Level: B2
Italian
Level: A2
SOFT SKILLS
INTERNATIONAL EXPOSURE + CULTURAL INTELLIGENCE (CQ)
Throughout my studies, I gained experience working in international teams and interacting with people from around the world. Moreover, thanks to my travels, I gained exposure to many cultures and explored differences in how people around the world live and think.
GROUPWORK ORIENTED
I gained vast experience working with other people to analyze data, deliver business projects, propose solutions and tackle problems – all in a collaborative environment.
OUT OF THE BOX SOLUTIONS & INNOVATIVE APPROACH
As my supervisors and professors can testify, I truly suggest and implement ideas from a perspective that no one would think of. Moreover, when it makes sense, I attempt to apply automation solutions and modeling programs to enhance my workflow,
Sections of my Portfolio
See different courses and specialization I completed on platforms, like edX and Coursera during my bachelor studies.
See photos from my travels, which contributed to my unique worldview and versatile predisposition.
Projects
Under Development
Explore quantitative trading models I developed and other data science projects I worked on.