Nathan Matharu's Projects
What I’m currently working on
Neuro-Kinesis (August 2025 - Present)
I’m working at Neuro-Kinesis, a medical device startup developing a diagnostic catheter with a superior signal-to-noise ratio and thereby greater resolution and fidelity.
Key Contributions:
- My primary role is functioning as the CAD specialist and maintaining engineering drawings. I’ve spent my time improving and automating parts of this process by creating a python package that directly communicates with SOLIDWORKS’s API and establishing the SOLIDWORKS PDM vault, importing the existing filebase into it.
- Authored a 10-page Standard Operating Procedure (SOP) for device assembly and packaging
- Diagnosed non-conforming PCBs through systematic troubleshooting, synthesizing electrical schematics and oscilloscope data to determine and document root cause for engineering review
- Deployed a SOLIDWORKS PDM vault with SQL Server backend to replace an ad hoc file system, migrating 150+ files
- Collaborated with an on-site aerospace development team to translate legacy hand-drawn schematic into 3D CAD models
Time-Series ML Pipeline for Financial Forecasting (January 2026 - Present)
Co-developing an algorithmic trading pipeline that ingests stock and options data to classify market periods into bull or bear regimes. The pipeline utilizes regime-specialized models (including LightGBM, LSTM, and RNNs) to forecast price movements and execute trades within a paper trading environment.
Working on such a project with a friend has been a rewarding experience. It has given me invaluable experience with messy, real-world data and with architecting codebases in a scalable, modular way with a custom Python package.
Construction (2017 - Present)
On weekends, I assist with electrical and management tasks at Taj Building Services, our family construction firm. This hands-on experience ranges from routing wires, testing, and diagnosing junctions to high-level project management. Navigating live construction sites has sharpened my ability to source materials efficiently, schedule cross-dependent tasks, and proactively minimize rework in the field.
Previous Projects
Semi-Soft Fingers (April 2021 - November 2025)
My undergraduate and graduate research involved spearheading the soft robotics branch of my lab. During this time, I learned the foundations of robotic kinematics, how to apply it to tendon-driven systems, and how to develop manufacturing techniques for silicone casting with functional embedded components.
This project iterated over three main designs: multi-durometer silicone, PVC-and-dowel, and silicone with 3D-printed bones.
See more on my publications page.
SC Solar Car Student Design Team (August 2021 - December 2023)
Over two years with the SC Solar Car Student Design Team, I transitioned from researching carbon fiber shell manufacturing to leading the mechanical design of the vehicle’s chassis. I conducted structural analysis via SOLIDWORKS FEA and collaborated cross-functionally to integrate the mechanical, electrical, and solar systems. Beyond the technical scope, this role was a masterclass in project management under tight constraints. Leading a volunteer team of undergraduate engineers required balancing aggressive project timelines with the realities of their academic schedules. This experience fundamentally shaped my approach to leadership, teaching me how to distribute critical knowledge, allocate resources dynamically, and drive progress through inspiration rather than formal authority.
Hand Tremor Mitigation Device (August - December 2023)
For the Senior Project of my Biomedical Engineering Bachelor’s, I co-designed and 3D printed a prototype to record and mitigate hand tremors with 3 of my peers. This idea originated from my work in my research lab, where I wondered if a cable-driven system could react and ameliorate tremors. This class challenged us to create this medical device from start to finish within a semester. We started with the MoSCoW method and market research, quickly realizing that any motors needed to run the system would likely be too heavy or too large to be comfortable for an end user. We pivoted our idea to using elastic bands and flex sensors to mitigate and record tremors, respectively. We programmed a state machine in LabVIEW and imported the data to Arduino for data analysis. Ultimately, the prototype was unable to successfully mitigate tremors in real-time. We discovered that without a dampening element, the elastic bands were just as likely to amplify tremors as they were to mitigate them. While the academic timeline restricted further mechanical pivots, this outcome provided a critical, hands-on lesson in system dynamics, the necessity of robust design inputs, and the importance of output verification