Tomasz Posłuszny

This is a marketing platform for structured financial products (gsmarkets.de).

My responsibilities included supporting the design and implementation of the microservices platform and building authentication and user management microservices, which enabled user experience personalization.

I helped build assets and prices streaming microservices to process and deliver hundreds of thousands of financial data messages within seconds. I have supported the implementation of the monitoring system (metrics, request tracing), which helps find anomalies and bugs quickly and improve performance.

I also supported the deployment configuration, which included packaging, package delivery, deployment scripts, network configuration, scaling, and database sharding.

Personal development project around (semi)automated trading:
• Back tester in C++
• Trading bot in C++
• Real-time data collector in C++
• ETL jobs in Python
• Data analysis and visualization in Python
• Scripts for managing "data lake" in S3

I implemented an image annotation tool for Windows from scratch. It was a Windows desktop application whose primary purpose was to facilitate the creation of a training dataset for supervised learning algorithms (such as neural networks). Technologies used: .NET framework, managed C++, and MVC pattern. Views and flows were implemented according to the UI and UX designs.

Publication:
While working on the project, I contributed to OpenCV and fixed the MOG2 algorithm implementation. The extraction process involved moving objects in videos.

Posłuszny, Tomasz, and Barbara Putz. "An Improved Extraction Process of Moving Objects' Silhouettes in Video Sequences." Advanced Mechatronics Solutions. Springer International Publishing, 2016. 57-65.
(See list of contributors here: https://opencv.org/opencv-3-0/)

I led the development of a Chromium-based internet browser and integrated an auto-update feature for both Windows and Mac using open-source projects.

Github.com/google/omaha
Github.com/omaha-consulting/omaha-server
Sparkle-project.org/

I designed and implemented a process for syncing with upstream Chromium major releases, developed and documented build and release processes, and implemented minor browser customizations and features.

An extended RocksDB Java API (RocksJava) related to options.
I added event listeners to RocksJava, which allows better monitoring and control of RocksDB internal operations. I also implemented multiple JNI benchmarks to optimize getting DB values via RocksJava, founding some approaches to be faster in certain scenarios.

As the primary developer, I built BrowseWiz, a browser extension that integrates AI directly into your workflow. I created this AI assistant to eliminate switching between research and separate AI tools, streamlining information processing and content creation in one tab.

Working independently (with UI design collaboration), I built:
• Page and video summarization for quick content understanding.
• Content generation for drafting alongside source material.
• AI automation with reusable prompts, custom instructions, multiple LLM support, and tool integration, creating a personalized agentic AI experience.

BrowseWiz runs via a non-intrusive side panel (Alt+B), keeping AI accessible without disrupting focus. This project shows my ability to develop a complete product independently, from concept to functional software with front-end, back-end, and API integrations. It delivers time savings for research, reading, writing, and AI agent management.