Bartosz Ptaszynski

At Vectra AI, I took on a POC solution and turned it into a scalable service as initial demand from customers and released the POC to early adopters, which was a great success.

As the senior developer on the still small team, I had to hit the ground running and dive deep into the maze of existing source code, automation, and infrastructure. Within the first three months, we managed to hit our significant milestone of removing dependency on the existing POC back end and replacing it with our new scalable solution.

The service itself is composed of a Linux service running alongside the detect software on physical and cloud servers, synchronizing data selectively with the cloud's back end used by security analysts. This work involved creating a new service, which integrates tightly with an existing large project and new cloud-based services supporting communication at scale.

This was a project for F-Secure to implement a back-end allowing to synchronize and backup content and contacts between the mobile applications and existing cloud storage and PC backup software.

The project was successfully deployed to the global user base through F-Secure partners, namely mobile carriers in France, Australia, and Mexico.

From initial design through implementation and setup of the infrastructure and CI/CD pipeline, this project was delivered mainly by myself with the aid of a web designer.

Various components in this solution include Arduino, C++ sensors, Go, ARM IoT edge gateway, Elixir with Phoenix queuing through MQTT and Inmarsat satellite modems, and an API with web portal. Deployed with GitLab CI and ArgoCD onto Kubernetes running on GCP. The solution had to be robust as the sensors were deployed at sea with the unreliable connections. Mobile 4G was used, falling back to satellite communications.

The back end by design had to be scalable and responsive, hence Elixir and deployment on Kubernetes.

Working in the SysOps team at Teamwork. I've introduced Kubernetes to the company and modernized monitoring in the company. During my time at Teamwork, I've also developed various tools to help automate and improve the visibility of areas of infrastructure before they had a chance to become an issue.

I've been part of a team working on a private cloud used internally by Fidelity. I have led efforts to improve our cloud's automation, security, and configuration management.

Automating and integrating F5 load balancing into C2C, external monitoring, and SSL certification. I've coordinated work with other teams on most of these efforts, helping with requirement gathering, documentation, and analysis.

Later, I joined the architects' team as an enterprise architect, aiding work on modernizing Fidelity's cloud infrastructure.

This Django project started with a very tight deadline of a month and was delivered to spec on time. The contract got extended to continue and implement new features. It's a CMS portal and a platform that allows wedding venue managers to promote venues and clients to make bookings. It also includes an embeddable booking calendar that hotels and venues can use on their websites.

Fabric is the next step from Zoom and Google Meet calls, where participants can share a virtual browser to make browsing the internet a more collaborative experience.

During my contract with Fabric, I introduced Kubernetes (AWS EKS) as a platform for the service, replacing a home-brewed scaling solution based on Lambda and AWS Auto Scaling groups. Moreover, I implemented a custom scheduler in Go on top of Kubernetes and ArgoCD, which allowed the deployment and management of the lifecycle of a very stateful WebRTC-based application. I improved the virtual touchpad scrolling inside the virtual browser and a custom Chrome extension running in a Docker container and streaming data via WebRTC. This involved debugging and coding in C++, Go, and TypeScript. Finally, I fixed several issues in infrastructure configuration, improving a few security issues in the process. This included applying the changes to infrastructure with Terraform and new infrastructure elements deployed via ArgoCD/Kubernetes.