Ņikita Širons

Software developer by week, climber on a weekend

Recent Projects

With probability map

In progress

2024 - Present

Expanded AWS infrastructure to EU for a consolidation project aimed at optimizing shipping processes by reducing the number of boxes per order.

2021 - 2024

Worked on system integration to enable multiple shipping carriers into the Amazon Seller Central platform, expanding the range of options available to third-party sellers and buyers while also improving convenience and cost-effectiveness (+10M packages shipped).
Developed and implemented a solution to correct shipping label errors, reducing address-related issues and cutting additional monthly costs by $200K, thereby improving operational efficiency.
Developed internal AWS microservices that facilitated the creation of a code analyzer and data pipeline, supporting automated analysis and data flow within the project.

2018 - 2019

Extended SDK by developing libraries that integrated seamlessly with other FEM software tools, enhancing software interoperability and expanding the functionality for end-users.
Partnered with clients to design and implement customized solutions, leveraging SDK and integrated libraries, resulting in the successful execution of multiple projects tailored to client needs.

2018-2021

Thesis: Radar-Lidar-Camera sensor fusion for 3D Object Detection. Worked on a study which aimed to improve the accuracy of 3D object detection in autonomous driving by fusing data from radar, lidar, and camera sensors. Given that each sensor has its own modalitites and different performance characteristics (like spatial resolution, field of view, etc.), it was an interesting challenge to find the optimal fusion strategy.
Teacher Assistant for the courses "Advanced Control" (2019) and "Linear Systems" (2020)

2015-2018

Thesis: Autonomous Race Drone. Awarded first prize for the final bachelor thesis, this project focused on designing a fully autonomous race drone to compete in and win the IROS 2018: Autonomous Drone Racing competition. My primary responsibilities included developing the lightweight computer vision software solution to detect gates and self-localize on the track and designing the hardware architecture.
DARE (Delft Aerospace Rocket Engineering). Member of the team, worked as a simulation engineer. I have worked on making web-based application for monitoring distributed simulation batches over multiple clients, developing parser to obtain aerodynamic coefficients of the rocket, making general post-processing tool and updating and testing dtst (DARE trajectory simulation tool).
Teacher Assistant for the courses "Programming & Scientific computing" (2017), "Exploring Aerospace Engineering project" (2018) and Mentor for first year students.