Projects
These projects show how Lucius applies machine learning, LLMs, robotics, and data infrastructure to practical problems in research, mobility, and finance.
The focus is on useful systems, clean execution, and measurable impact.
Robotics
Architected an AI-powered robotics platform integrating Boston Dynamics Spot, NVIDIA Jetson Nano, OpenAI GPT-4, and a cloud-hosted REST API to enable autonomous campus navigation and conversational AI interactions.
Developed secure robot-control infrastructure using GraphNav, ROS2, encrypted VPN networking (Tailscale), token-based authentication, live telemetry, and real-time camera streaming for remote operation.
View the live project →
Quantitative Analytics platform
Built a scalable quantitative analytics platform that enables investors to analyze and optimize stock portfolios using machine learning models, risk metrics, and financial simulations.
Designed machine learning infrastructure supporting both real-time and batch financial prediction workloads.
View on GitHub →
Chrome Extension
A beautiful 3D network visualization Chrome extension that displays LinkedIn connections as an interactive graph with advanced search, pathfinding, dragging, and camera controls.
Built with Three.js, Chrome Extension Manifest V3, and bundled local modules for performant, CSP-compliant rendering.
Installation instructions are in the repository.
Research
Developed AI-powered automation solutions for research administration workflows, including document classification, workflow routing, information extraction, and decision-support systems.
Still under construction.
Navigation