Cai Lin

Stock Intelligence Pipeline

• Tools: Python, AWS (EC2, Bedrock, SES, S3, EventBridge), yfinance, Jekyll, GitHub Actions
• Screens 2,700+ US stocks daily using NASDAQ API with S&P 500 priority coverage and yfinance technicals
• Claude (via AWS Bedrock) generates daily briefs with top picks, fibonacci entry/exit targets, and market direction calls
• Paper trading simulator auto-manages positions with stop losses and profit targets, tracking real P&L
• Short screener identifies bearish setups with entry/cover/stop levels validated by Claude
• Fully automated EC2 pipeline: screener → analysis → email → blog publish via GitHub Pages

Sports Betting Analysis

• Tools: Python, AWS Bedrock, ESPN API, BettingPros, Playwright, Jekyll, GitHub Actions
• Fetches player props from BettingPros + Underdog Fantasy, projects stats from ESPN game logs
• Quantitative edge scoring with Claude validation and 50/50 OVER/UNDER balance enforcement
• Smart scheduling: checks ESPN scoreboard daily, auto-skips offseason sports
• Daily email reports + automated blog publish via GitHub Pages

Job Application Tracker

• Tools: Google Apps Script, Gmail API, Google Sheets, clasp CLI
• Scans Gmail every 5 minutes for application confirmations and rejection emails
• Extracts company name, role, and source (LinkedIn, Indeed, Greenhouse, Lever, etc.) using regex pattern matching
• Detects rejections in threaded replies with fuzzy company matching across multiple roles
• One-click template setup for other users; Notion database integration available

Previous: J.P. Morgan Virtual Experience

• Tools: Python, TypeScript, JavaScript, HTML/CSS, Perspective, node.js, Chart.js, pip, npm, Pytest, Git/GitHub
• Developed dynamic and interactive data visualization dashboards to support JPMorgan Chase traders by leveraging Python, TypeScript, JavaScript, HTML/CSS, Perspective, node.js, Chart.js
• Implemented new features for the Perspective dashboard, enhancing functionality and usability for traders and stakeholders, resulting in a projected 25% increase in informed decision-making accuracy within financial markets
• Engineered a module to retrieve and process financial data feeds for two historically correlated stocks, enabling real-time monitoring of price movements with an accuracy rate of 95%
• Orchestrated custom charting components within the Perspective framework to visualize the correlation between the two stocks over a span of up to 10 years, enabling traders to identify instances of divergence from historical norms with a projected precision of 90%
• Rectified broken TypeScript files within the repository through patch updates, ensuring the proper functioning of the web application and correct output of data visualizations for traders, resulting in a 100% improvement in application stability and usability
• Implemented unit testing with pytest for patch updates, resulting in a 50% reduction in post-update issues and a 20% increase in overall code stability
• Conducted performance optimization and scalability testing of Perspective dashboards
• Integrated the adjusted data set into Perspective, leveraging its capabilities to visualize live and historical data feeds in a clear and intuitive manner
• Developed proficiency in setting up a development environment by installing Python, forking and cloning the starter repository, and installing project dependencies, enabling efficient collaboration and code development
• Familiarized with engineering tickets and their role in project management and task allocation within the development team

Stock Price Prediction

• Tools: Python, Pandas, Scikit-learn, Jupyter Notebooks, Matplotlib, Seaborn, Numpy, Yahoo Finance API, Git/Github, Streamlit, Support Vector Regression Model, LSTM Model
• Developed a comprehensive stock price prediction project leveraging Python, Pandas, Scikit-learn, Jupyter Notebooks, Matplotlib, Seaborn, NumPy, and the Yahoo Finance API
• Deployed the web application on cloud platforms like Streamlit Sharing, enabling users to access the prediction tool and sentiment analysis tool from anywhere, resulting in a 50% increase in user engagement
• Implemented machine learning models such as Support Vector Regression and LSTM to forecast stock prices with an average accuracy of 85%
• Engineered interactive visualizations and dashboards using Matplotlib, Seaborn, and Streamlit, enabling users to explore over 5 years of historical stock data, model predictions, and sentiment analysis results with real-time updates
• Implemented user-friendly features such as dropdown menus, sliders, and interactive plots, resulting in a 30% increase in user engagement and retention
• Utilized Git/GitHub for version control, code review, and project management to ensure an organized workflow and efficient code management throughout the development process
• Conducted sentiment analysis using Twitter feeds to gauge market sentiment and its impact on stock prices, employing Natural Language Processing techniques and sentiment analysis libraries
• Analyzed sentiment on Twitter feeds related to stock sentiment, extracting data from Kaggle datasets, and amalgamated it with Yahoo Finance data of the stock to create a unified dataset for analysis
• Processed and analyzed large datasets containing historical stock prices and Twitter feeds, achieving a data cleaning accuracy of over 95% and reducing data preprocessing time by 30%
• Utilized Python libraries such as NLTK, TextBlob, and VADER for sentiment analysis, processing text data to derive sentiment scores and trends
• Integrated BERT (Bidirectional Encoder Representations from Transformers) sentiment analysis into the project pipeline, leveraging its advanced natural language processing capabilities to analyze Twitter feeds for sentiment trends. This analysis was then used to inform trading decisions over a 5-year period, demonstrating the effectiveness of incorporating state-of-the-art NLP techniques in stock trading strategies
• Executed a trading strategy leveraging sentiment analysis, yielding a 15% higher return compared to a traditional buy-and-hold approach over a 5-year period

ArtifyMe

• Tools: Python, NumPy, SciPy, OpenCV, Streamlit, HTML/CSS, JavaScript, Git/Github, Jupyter Notebooks, Convolutional Neural Networks, Generative Adversarial Networks
• Developed "ArtifyMe," a Image-to-Drawing Image Recognition Web App that allows users to upload images and transform them into cartoon or anime like drawing
• Utilized pre-trained deep learning models including Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs) for image detection and recognition tasks, achieving a projected accuracy rate of 90%
• Used image processing algorithms for encoding, decoding, feature detection, image matching, segmentation, and transformation using the OpenCV library, resulting in a projected 95% success rate in image transformation
• Introduced an image uploading feature allowing users to upload their selfies directly into the web application for transformation into artwork
• Added a feature allowing users to choose to convert their uploaded images into anime or cartoon-style artwork using advanced image processing libraries from CartoonGAN and AnimeGAN
• Implemented customized filter and image adjustment features, resulting in a 50% increase in user engagement
• Integrated user-friendly interface features, including scrollbars and image adjustment options, resulting in a 30% increase in user engagement and a 20% improvement in overall user satisfaction
• Leveraged Streamlit Cloud platform for deployment, ensuring seamless accessibility and scalability of the application
• Conducted rigorous testing and validation to ensure accuracy, reliability, and performance of the image recognition and transformation algorithms
• Utilized Git and GitHub for version control, ensuring efficient collaboration and tracking of project changes

City University of New York

Awards & Certifications

• CompTIA ITF+
• AWS Certified Cloud Practitioner
• JPMorgan Chase - Software Engineering Simulation
• JPMorgan Chase - Software Engineering lit Simulation
• JPMorgan Chase - Agile Simulation
• Meta - Introduction to Back-End Development
• Meta - Version Control
• Meta - Programming in Python
• Meta - Introduction to Databases for Back-End Development
• Educative - Grokking the Coding Interview
• 2nd place in Google Tech Challenge 2019