About Me
Hello! I’m Batuhan, a software engineer at CME Group and a graduate of University of Illinois at Urbana-Champaign with degrees in Computer Science and Economics (May 2024). I’m passionate about applying technology to solve challenging problems, from trading infrastructure at CME to innovative personal projects.
At CME, I’m currently working on Java-based real-time risk management systems.
One project I’m especially proud of is my algorithmic trading system built for the Solana blockchain. Harnessing machine learning, real-time blockchain data, and optimized infrastructure, this system effectively predicted new token price movements and achieved significant returns during peak trading periods (1500% ROI in one week).
I’m always eager to tackle deep technological challenges, especially in fintech, and AI-driven solutions. Feel free to contact me at [email protected] or connect via LinkedIn. Looking forward to connecting!
Experience
- First Rotation (June 2024 - January 2025):
- Worked on a full-stack application based on Backstage, an open-source software catalog for building developer portals
- Designed and implemented custom integrations with internal services, streamlining developer workflows and significantly reducing friction in software discovery and deployment.
- Second Rotation (January 2025 - Present):
- Working on Java-based real-time risk management systems using Spring Boot and Kafka
- Using in-memory databases for live risk / margin calculations
- Summer 2023:
- Designed dynamic validators for trading infrastructure using Java, enabling seamless configurations and revolutionizing software testing flexibility
- Reduced execution time by 85%, significantly improving software development and QA processes
- Enhanced early error detection, significantly saving developer hours previously spent on troubleshooting
- Transformed the testing tool, positioning it as a vital component in CME’s routine software development and QA processes
- Summer 2022:
- Refined a critical Java application, streamlining real-time testing for CME’s buyer-seller matching engines
- Implemented a comparative analysis feature, allowing for real-time contrast between established outputs and new prototype versions, enhancing quality checks
- Implemented automated alerts via Splunk, reducing manual checks and increasing system reliability
- Leveraged Kafka for real-time streaming, reducing system downtime and enhancing reliability metrics
Education
University of Illinois at Urbana-Champaign
Computer Science + Economics
January 2022 - Present (Expected Graduation: May 2024)
Major GPA: 3.93
Purdue University
Computer Science
August 2020 - December 2021
GPA: 3.56
Projects
Solana Trading System | Python, Node.js, sklearn, PostgreSQL, AWS
https://coinstats.app/address/BgrVfkzUTpKMb3M4A5HWgksECZ7FrkRHaj3ofhmDxRxp/
- Developed a machine learning-powered trading system utilizing multiple ML models to predict newly launched solana crypto price movements
- Achieved returns of 1500% ($50 to $800) within 1 week during peak performance, with >1000 coins traded
- Deployed and maintained a dedicated Solana validator node to optimize trade execution speed and minimize latency
- Implemented data processing pipelines integrating blockchain data, social media metrics, website info (such as whois & technologies used) and trading patterns
- Engineered an AI-powered chatbot using GPT-4, enabling automated, intelligent customer support for Shopify merchants.
- Utilized vector databases to index store pages and products, enabling RetailGPT to retrieve and present relevant information upon customer queries.
- Developed a scalable backend with Node.js and Express, deployed and managed on AWS.
- Crafted a responsive frontend using React and JavaScript, focused on providing an intuitive and seamless user experience.
- Designed and developed a cryptocurrency trading bot that identifies and capitalizes on pricing discrepancies between various exchanges.
- Engineered the bot to connect to multiple exchanges through their APIs, fetch real-time price data, and analyze for profitable trading opportunities.
- Implemented an automated trading system that executes trades instantly when profitable discrepancies are detected, based on pre-defined criteria.
- Open-sourced the project on GitHub, where it has garnered significant community engagement with 160+ stars and 60+ forks, indicating its utility and popularity among developers.
- Developed an automated email-to-audio transcription service.
- Utilized Cloudflare and Google Cloud Platform for scalable email handling and audio generation.
- Implemented personalized private RSS feeds, enabling users to securely access audio transcriptions directly in their podcast apps.
- Leveraged LLMs to clean and enhance transcriptions, structuring them into well-formatted chapters with relevant headers.
- Developed a web tool helping users find and compare protein powders by nutritional value, price-per-gram, and dietary filters.
- Built automated Amazon data extraction pipelines leveraging multimodal LLMs to pull accurate, detailed nutritional information from product images.
- Designed structured data system supporting product variations and intuitive sorting/filtering options (protein per dollar, calories per protein, dietary restrictions, user ratings).
- Engineered a tool to detect reversal points on stock prices by identifying clusters of local min and max points
- Utilized Yahoo Finance for stock price history and visualized charts with marked reversal points.
- Project open-sourced on GitHub, garnering 290+ stars and 50+ forks.
YoutubeTimestamps | JS
https://chromewebstore.google.com/detail/youtube-timestamped-comme/hemochpgdbblobafehgcngacdjmaledg- Built a Chrome extension enhancing YouTube videos by showing timestamped comments on the timeline.
- Parsed timestamps from video descriptions and comments dynamically.
- Added interactive markers and tooltips for seamless user navigation.
- Developed a responsive and intuitive user interface using React and JavaScript, empowering app developers to track and compare their app’s performance against competitors across Google Play and Apple App Stores.
- Implemented real-time review tracking feature for the Play Store, enabling developers to monitor user sentiment and feedback over time, and correlating this with specific updates to their app.
- Designed and implemented a Django backend, serving as a RESTful API that scrapes, processes, and delivers app performance data to the frontend.
- Developed a comprehensive GUI application enabling seamless cryptocurrency trading.
- Engineered versatile trading features, including arbitrage detection, market making algorithms, back-testing capabilities, live trading of Technical Analysis Strategies, dollar-cost averaging implementations, and trailing stop functionalities.
- Utilized the Xchange library to facilitate support for dozens of leading cryptocurrency exchanges, simplifying multi-platform trading for users.
- Developed “iDetect,” a real-time emotion analysis application, at the iMAKE hackathon, enabling speakers in virtual meetings to gauge audience engagement by analyzing video streams.
- Utilized Python’s DeepFace library for facial recognition and emotion classification
- Designed and implemented data visualization components to display real-time emotion statistics via tables and graphs.
- Coded a tool to manage large spreadsheet files due to the limitation of ~1 million rows in excel files.
- Can split a larger file into many sheets or subfiles of various file types, adjusts formulas to point to new cells.
- Supports different file types (.csv, .xlsx, .xls) and has many customizable settings.
- Regularly searches the RMV website for desired road test appointments
- Notifies the user when an appointment is found
- Chrome extension that allows you to take notes as you are watching a video.
- Can later on click on a note which opens up the video at the right timestamp.
- Done during ‘Hello World 2020’ hackathon at Purdue.
- Created “Spotify Song Announcer,” a Python application that detects song changes in Spotify and audibly announces the song name and artist, enhancing user experience.
- Utilized SwSpotify for song information retrieval, bypassing the need for Spotify API and enabling offline usage.
- Integrated pyttsx3 for text-to-speech conversion.
- Configurable options for voice selection, announcement delay, and speaking rate, allowing for personalized user settings via command line arguments.