Evgeny Orlov

Part of my responsibilities in my last place of work was to create algorithms to determine the activity of underground loaders on mining sites. The raw input from Machine Vision told us what state and what prediction level it sees, and the goal was to create algorithms that can decide what each loader is doing. I invented and implemented algorithms within a month, including signal filtering logic and special state machine. During the deployment on-site, they were tuned and provided much greater accuracy than human reports, which is a fantastic result. Besides that, I completed a lot of refactoring in the 15-year-old legacy code to make it controllable to minimize the number of issues and time spent, including the creation of a data access layer with EF-style migrations, suitable for MsSQL and SQLite with T-SQL syntax converter to unify SQL queries for both databases, unified the communication layer, based on gRPC, created production logs, automated logs delivery from the mining site units onto the server, and many other activities. Because of my activity, the company was able to deploy the solution on several large mining sites quickly and resolve all major issues in time.

After half a year of working at Visitech in .NET 4.6 and an Entity framework back-end, I was tasked with developing the Node.js back-end where I had my first Node.js development experience. The only available option to work with databases was the Sequelize package. I built the back-end template together with a micro-service. To the basic Sequelize functionality, I added Entity framework-style code, first migrations and recursive algorithms to CRUD, and a graph of linked objects into and from the PostgreSQL database using simple mask syntax to control which parts of the object graph should be affected or obtained. In other words, the front-end sent raw JSON data containing a graph of linked objects, and with the help of my add-on, it required no data transformations to CRUD it into the database all at once. I also completed many other tasks such as converting hard-coded business logic controlled by a JSON configuration, applied SSO to the product with Node.js micro-services, made simplifications to achieve a better developing speed, and added implemented JWT tokens to Node.js and C# micro-services. All that enabled the company to continue further development with minimal resources not go under.

I spent 11 years developing and supporting a complete set of banking applications to accept and process payments from users to final providers such as mobile or home internet providers. I created a client-server solution with many integrations with providers and other payment systems and ATM software with many interactive UI skins to accept payments. I integrated these with many ATM hardware (validators, printers, dispensers, watchdogs, plastic cards) and created a payment module used directly from cash-desks, wrote reports, prepared about 50 thousand provider templates, including designing and converting their logos, and related tasks. I supported clients with a 24/7/365 schedule for many years to help them avoid long pauses in payment processing. Many challenging tasks were completed during the period, for example, processing payments through a site for users (not a payment gateway) or using SMS-sending services to provide two-factor authentication for users to access their personal accounts on an ATM. I have an author's certificate for ATM and payment module software as I am the only author and owner of the software.

This was a 3D action game about patrol torpedo boats during WW2, where I was the lead programmer. In the 2003-2004 year, graphic adapters were not as good in processing pixel shaders, and it had been decided to create geometric common and dynamic waves on the sea. It was quite a challenging task, as there were hard limits on the number of polygons and the sea required to have a set of LODs. I think the result is still impressive; you can check it by downloading the video. I created all particles such as clouds, fire, and dynamic fog in the project and many other parts, including audio support for ship's devastation algorithms and others.

In early 2000, a designer, musician, and I created fantastic software for music composing called Samplisizer. We spent 1.5 years developing it, and another 1.5 were trying to sell it. The app is excellent even now. You can still download its trial version from many sources and works in modern Windows versions. I was the only programmer and implemented cool features such as adaptive app borders to make it look great regardless of the background and effects such as fire, blur, and others. To improve performance, it has a pre-render feature. All loops in the app consist of sounds and can be used with any BPM. It is still quite a cool piece of software, considering it was created in 2002. It used to be the fourth in Google search in the music sampler app section without paying for promotion.