Engineering Projects
Selected Projects
Systems built at the intersection of AI, robotics, automation and data engineering.
Hybrid Adaptive BCI
Problem
Manual control is impractical for users with severe motor impairment, and in industrial settings where hands-free operation is needed.
Build
M.Sc. thesis engineering a hybrid adaptive Brain-Computer Interface that decodes real-time EEG with deep learning to control industrial robotic manipulators and medical assistive devices.
Technology
AI Job Agent
Problem
Manually searching, filtering and evaluating job postings across multiple sources doesn't scale.
Build
An engineering system that combines AI agents, LLM analysis and workflow automation to collect, process and act on job data end-to-end.
Technology
ROS2 Robotics Systems
Problem
Bridging AI-driven decision-making with real robot behavior requires a dependable control and simulation layer.
Build
Robot control and simulation work built on ROS2, exploring intelligent behaviors and human-robot interaction.
Technology
Intelligent Data Pipeline
Problem
Intelligent systems are only as reliable as the data feeding them — inconsistent pipelines undermine everything built on top.
Build
A data engineering system for ingesting, processing and serving structured data through reliable, intelligent pipelines.
Technology
Jet Engine Design and Performance Simulation
Problem
Characterizing a jet engine's efficiency and fuel consumption across operating conditions requires validating a design before physical prototyping.
Build
B.Sc. graduation project designing a complete jet engine assembly — compressor, combustion chamber, turbine and exhaust — using thermodynamic cycle analysis and fluid mechanics, modelled in CAD and validated with ANSYS and MATLAB simulations.
Technology
AI Automation Systems
AI Automation Systems
Intelligent workflows that combine LLMs, AI Agents, APIs, databases, n8n and automation into engineered systems — not one-off scripts.
AI Job Agent
An engineering system that continuously collects job data, analyzes it with an LLM-driven agent, and delivers structured, actionable results — designed as a full pipeline, not merely an n8n workflow.