Pavle Jonoski

An open-source, scalable, and ready-for-use microservice framework. The company I worked for used the framework successfully on multiple projects for different clients. In addition to the software architecture for this framework, I implemented the base microservices, created libraries for the framework in various languages, and wrote deployment scripts for Kubernetes and Docker Swarm. You can see the framework repositories at https://github.com/microkubes.

A data management cloud platform that offers CKAN as a service in a SaaS model (https://amplusdata.io). I co-developed the software architecture (based on microservice architecture), played a key role in implementing many of the services, and supervised other developers and reviewed their work.

FOCUS AREAS
* Developed the platform design/architecture
* Designed and implemented modules/services of the platform
* Wrote deployment scripts for Kubernetes with Helm
* Solved tough-to-crack development problems
* Guided developers through the product development process
* Reviewed code
* Fixed bugs

A specialized search engine that is optimized for specific languages that gathered data from the wild; in other words, from multiple domains and not just scraping the data from specific sites. As a result, the search engine had to handle the discovery of new domains and sites, proper data extraction, and crawler traps.

I designed the distributed system from the ground up, on top of which the engine was implemented. The distributed system executes both real-time and batch jobs. It was designed to handle large amounts of data, function autonomously, and scale horizontally to hundreds (if not thousands) of nodes.

To solve some of the problems that arise—such as ranking, job scheduling, detection of traps or tar pits, performance, and efficiency estimations—for which there is no off-the-shelf solution, I had to implement solutions based on academic papers, perform simulations, and create mathematical models to calculate some behaviors of the system. Data analysis and machine learning techniques were necessary to solve some of the problems.

I implemented the system's core and then led a team of engineers to implement the full system. I also performed code reviews and assisted with the project management methodology.

A lightweight log aggregator that collects logs from multiple machines and provides the aggregated logs on a single endpoint. It offers real-time viewing of the logs (a log stream) or a historical search of the logs. This is the project that I use as a tool during development, especially when I need to track multiple processes or have many instances on different machines and/or containers. It has collectors (agents that watch and collect the log lines) in Python and Go. I did the full design and implementation.