Mladen Krstic

I participated in the Palladium 2020 algorithmic challenge organized by Codility and sponsored by Iress (https://www.iress.com) and I won the Golden Award. I also earned a hall-of-fame status because I was among the first three participants who submitted the flawless solution.

Statistics:
https://app.codility.com/programmers/challenges/palladium2020/

I participated in the Rhodium 2019 algorithmic challenge organized by Codility and I won the Golden Award.

Statistics:
https://app.codility.com/programmers/challenges/rhodium2019/

I participated in the Ruthenium 2019 algorithmic challenge that was organized by Codility and I won the Golden Award.

Statistics:
https://app.codility.com/programmers/challenges/ruthenium2019/

Swiss Re, one of the world's leading insurance and reinsurance providers, created a B2B2C platform called iptiQ, allowing other insurance companies to use this comprehensive solution to extend their business processes. It is a scalable, microservices-based, and event-driven platform that covers all aspects of the insurance business. Its efficiency is confirmed by the fact that Zurich Insurance Group, another leader in the industry, is actively using it.

During my engagement on this platform, I worked on different system segments and became familiar with the elaborate business processes previously unknown to me. I performed significant roles in delivering complex business and infrastructure-oriented functionalities that allowed me to get detailed insights into what an end-to-end lifecycle of delivering a product looks like in a complex distributed system for such a large company. These included business specification to implementation, testing, deploying, monitoring, and tracing in different environments.

Airmeet is a platform for hosting virtual conferences. The idea behind the platform is to allow organizers to schedule not only individual interactive sessions but to simulate complete events from registering attendees, creating sessions, registering speakers and hosts, allowing participants to interact in a social lounge between the sessions, registering booths and sponsors, and more. Starting as a startup, they quickly raised 15 million dollars from investors and grew up to a company with hundreds of employees whose platform is used by some large clients such as Volvo, Philips, Walmart.

During this engagement, I had a great experience and the opportunity to do a lot of stuff, from research and analysis to designing and implementing the solutions multiple times, taking a leading role for the parts that I was working on. I am especially proud of the Polls API that I designed in coordination with other teams. The API handles the voting process of the attendees in the session, and it can handle jumps from low traffic to sudden bursts of up to 10 thousand requests per second. At the same time, aggregated data about voting results are pushed in real time to the clients no matter how intensive the traffic is.

Moneyhub platform is one of the first global Open Banking platforms. It offers different kinds of services to the customers that can be individuals as well as companies. Customers can manage and take control of their finances through the web and mobile applications as well as secure APIs. Platform connects to different services accounts such as banks, cards, pensions, savings and investments.

Technical infrastructure is based on Node.js microservices and Docker environment. I was involved in different important processes as a full stack developer and I had an opportunity to develop and to upgrade features that are spread throughout the multiple layers within this complex system. Using cutting-edge technologies and best practices I had an opportunity to take responsibility for features that handle high throughput, to optimize reports with huge data sources, to refactor APIs in order to support easier upgrading and to cover flows with unit and integration tests… I was also involved in discussing and actively proposing technical solutions and ideas.

In coordination with the data science team, I developed an app that allows economists to prepare data and generate different types of economic reports along with automation of the different flows.

Data preparation is performed monthly, and different paragraphs that can be automatically resolved for every month using the data from the different data sources. This automatic process gives an economist an initial template for every paragraph and the ability to additionally manually edit those paragraphs.

With about 200 variable parts in various paragraphs, the challenge was resolving all of them in a consistent way. Another challenge was to optimize the resolving speed of all those paragraphs before they appear on a page for further editing.

After implementing those key mechanisms, I used them to handle the entire full-stack development of all the features that give economists a complete flow and allows them to effectively finish the job from a user perspective.

Using Node.js, GraphQL, and React as a basis in technical terms, I also had an interesting experience in terms of the integration with different storage types such as PostgreSQL and Contentful.

I handled the full-stack development of a platform that allows the collection of marketing data. The app can be embedded in external websites and users can automatically share results on different social networks.

My role was to develop new and upgrade existing functionalities using AWS serverless architecture on the back end and using AngularJS and React on the front end.

This is an application that allows users to implement the automatic tracking of sports events results by fetching external feeds or to track results manually if the specific sports event requires more detailed tracking.

Users also have options to set the level of details for tracking a specific sports event or competition—to exclude a sports event or competition from tracking and to retroactively control and edit the result of the ended events. Not only is it possible to track the results of the matches, but it is possible to edit the details of each score (goal, offside, foul, point, rebound, and more) individually (the time when it happened, identify the player who scored, and more). With many other features and details that make the everyday work of a user easier, this application is one of the key business mechanisms of Mozzart's, a betting company that is operating in the market with tens of millions of users.

Moving a result-tracking functionality from the old information system to the separated application was a part of the process of migrating the entire information system from monolithic to microservice architecture. I separated the full-stack application that didn't have any value for the company's business while it was not integrated with the rest of the system. So the other important projects, besides the separated application itself, were modules for the integration with other functionalities.

There was a module of the legacy system which processed tracked sports events concurrently in real-time and actually simulated results tracking in the legacy system itself. My main role was developing the adapter between the new results tracking application and the functionalities in the old system and to ensure that it functioned reliably in production. I also played a crucial role in the implementation of the microservice for the management of Teletext pages.

A Docker containerization platform is a strategic component in planning the migration to microservices for the betting company, Mozzart. One of the longterm plans was to apply Docker not only as a production platform but also as a platform used for every local development effort instead of using Vagrant. I actively researched Docker capabilities and moved the key infrastructure components (Oracle databases, MongoDB databases, ActiveMQ brokers, and more) from Vagrant to Docker. As a result, Vagrant was not used anymore in Mozzart and Docker is generally accepted within the company's IT sector and developers have not only minimized the use of hardware resources but also have a much easier and practical way of setting up local environments.

In order to develop a centralized front-end dashboard app to be used for different jobs within the Mozzart, AngularJS was chosen as the mainstream tech. Since the AngularJS hadn't been used at Mozzart, I needed to research it, to organize presentations and tutorials for other employees, and to set up the main structure of the mentioned dashboard app. As a result, both AngularJS and the dashboard app are now extensively used in the company.

PetrolSoft is an IT company that offers ERP solutions to a few gas station chains. Their system was already in use by a few large clients and it was designed to be distributed with local servers that are integrated with a centralized server, all via a common API so that a centralized solution should be replaceable.

Nestro Petrol, a large gas stations chain owned by Zarubezhneft looked for such a solution and RetailConnect, the ERP software developed by CBS Systems, was accepted as the most suitable solution for the above-mentioned centralized server.

I was working for CBS Systems at the time so I was responsible for developing the integration between RetailConnect and PetrolSoft's distributed system and adapting RetailConnect for the specific needs of the client. Every bookkeeping document, invoice, or any new item in the assortment must be synchronized through the integration module that I developed. I also had the opportunity to work with different technologies on different layers and to achieve a consistent state between Microsoft SQL Server as a middle layer and PostgreSQL for storage.