Carter Tate

Cornell CS • Seeking SWE / Data / ML roles

Carter Tate

I build software systems that turn messy data into usable predictions, recently developing ML-driven tools for revenue modeling and optimization.

EmailLinkedInResume
Carter Tate headshot

About Me

I’m a Cornell Computer Science student focused on building data-driven products and ML systems. My recent coursework that interests me most is related to Machine Learning, Operating Systems, AI Agents and Rationality, and Software Engineering.

Outside of academics and projects, I spend much of my time contributing to the Cornell wrestling team. Competing has taught me to be comfortable with high performance team dynamics, operaing under pressure, and improving quickly over time with feedback loops.

Cornell Wrestling team
Carter Tate wrestling match

In my free time, I'm interested in Formula 1 (strategy, engineering tradeoffs, and marginal gains) and different branches of philosophy that I can learn from to better reason about the world around me.

Formula One racing
Marcus Aurelius bust

Internship Experience

IGT logo

International Game Technology (IGT)

Core Studios Technical Intern · June 2025 – August 2025

  • Developed and trained deep neural networks (scikit-learn), XGBoost, and Random Forest models to predict slot-machine session dynamics, focusing on behavioral modeling and extensive custom feature engineering for high-dimensional gameplay data.
  • Analyzed SHAP values to identify key drivers of longer player sessions and increasing bet behavior, and produced confusion matrices to rigorously evaluate predictive accuracy.
  • Findings directly informed strategies for real-time slot machine game math adjustments at a global leader in the gaming industry with approximately $2.5B in annual revenue.
  • Built automation scripts and testing frameworks in Unity to reduce manual debugging time and improve developer iteration speed within the Core Studios pipeline.

Projects

A few things I've been up to recently.

Casino Coin-In Modeling

Predict monthly coin-in using engineered spatial features and ML models.

  • Built a supervised ML pipeline with feature engineering on floor layout + traffic proxies.
  • Compared models (XGBoost, RF, LightGBM) and utilized cross-validation to analyze performance
  • Used SHAP to explain drivers of predicted revenue and guide optimization.
Pythonscikit-learnLightGBMRandom ForestSHAPOptimizationFeature EngineeringClustering
Deep-dive with Visuals!

Air Traffic Control Simulator

Built an interactive air-traffic control simulator in OCaml with a live radar UI and a programmable command console for real-time aircraft control.

  • Developed the backend simulation engine to model aircraft motion in a 2D airspace and maintain consistent state over time.
  • Integrated backend ↔ UI communication so radar updates, command log events, compass heading, and runway indicators stayed synchronized in real time.
  • Implemented a programmable console supporting commands like changing headings, spawning aircraft, and removing aircraft during live simulation.
  • Collaborated in a 4-person Agile team (sprints, task breakdown, iterative demos) with clear ownership boundaries and frequent integration.
OCamlAgilesimulationstate managementUI integration

Huffman Compression Program

Built a Huffman file compression system in C using priority structures and Huffman tree encoding for efficient lossless compression.

  • Implemented a Huffman file compression system in C by building and traversing a Huffman tree.
  • Used priority stack/queue structures to construct codes and encode characters into variable-length bit patterns based on frequency.
  • Produced correctly formatted compressed output to achieve efficient lossless compression.
Cdata structuresHuffman codingfile I/O

Intelligent Image Segmentation Tool

Created a Java intelligent scissors image editor using pixel-graph modeling and Dijkstra’s algorithm for real-time selections.

  • Built a Java-based intelligent scissors image editor application with a Swing UI.
  • Modeled images as pixel graphs and applied Dijkstra’s algorithm to compute optimal selection paths.
  • Delivered a real-time selection tool with visual progress indicators for interactive image editing.
JavaSwinggraphsDijkstra’s algorithm

Optimized Matrix Multiplication for Performance

Implemented and benchmarked cache-aware matrix multiplication techniques in C to improve performance via locality and tiling.

  • Implemented and benchmarked cache-aware matrix multiplication algorithms in C (tiling, memory layout, access stride optimizations).
  • Focused on spatial locality and instruction-level optimization to reduce cache misses and improve throughput.
  • Analyzed runtime across implementations and quantified performance gains using profiling tools.
Cperformance optimizationcache localityprofiling

Contact Me

Explore more here:

EmailLinkedInResume