Denis Rechkunov Email & Phone Number

Berlin, Germany

Contributing to the OTel future of Beats.

Berlin, Germany

Working on Elastic Agent and Beats in the Ingest Data Collection team with a main focus on file ingestion (Go, Elasticsearch, Kibana, ESS).Among my notable contributions are:* Introduced the golangci linter to our repositories (initial configuration, team alignment on all rules, guidelines, etc).* Created the Filebeat registry analyser tool that enabled our support to easily diagnose known issues.* Added an automated migration flow from "log" input to "filestream", also known as the "take over" mode, that made it easier for our customers to migrate from the deprecated "log" input type.* Added a new file identification mechanism based on fingerprinting and wrote a blog post about it. This solved a lot of data loss and data duplication cases for our customers.* Actively participated in benchmarking and performance optimisations.* Was responsible for CI and security-related topics in our team.

Berlin, Germany

Worked on the platform team, mainly with Go, Kubernetes and PostgreSQL.My team was responsible for building services that provide:* Data/Asset Catalog for large organizations to track their data landscape: search, tags, metadata, links, glossary etc.* IdP — identity provider for the whole platform: authentication, authorization, user management, permission management.* Data Store — a service that enables uploading data in formats like CSV and integrating it in relational database models, so users are able to combine them with other relational databases using our Spark-based query engine.* PGQL — a service that allows access to any SQL-based data source via a GraphQL endpoint.* Queue and job system.* Notification system.* Open API spec‐first code generator that we've built our workflow around.

Berlin, Germany

Initially worked on the RESTful API that powers the company's mobile apps (Scala, Play Framework).Then moved to another team and worked on the API gateway for external partners which is an internally used fork of Skipper (Go). This gateway solved authorization, throttling, routing, circuit breakers and other common problems that usually occur on the ingress level.

Berlin, Germany

First worked on a set of framework agnostic components. Adopting the idea of web‐components we created our own runtime library for using components with several frameworks like Angular, React and GWT. It was used for building AWS consoles using consistent and design guidelines‐compliant UI‐components.Then worked on the static website deployment flow in AWS (TypeScript/React). This enabled AWS customers to easily deploy a static web-site seamlessly using underlying AWS services like Route 53, S3 and CloudFront.After that switched to some back-end work and helped to create a platform for running AWS consoles (Java, Spring). This solved common integration issues faced by internal developers when adding new services to the AWS family.

Novosibirsk, Russia

Leading front-end development on the Flamp team:* Adapting the web-component framework (Catberry.js) to the team's needs.* Hiring and mentoring new team members.* Presenting team achievements.

Novosibirsk, Russia

Working as a part of the Flamp team (review service). I was responsible for building a brand new version of the main web‐site which was supposed to become an isomorphic web‐application. This means the web-app would consist of 2 parts: a back-end that renders HTML for search crawlers and a front-end that runs a single page application in the browser without reloading the page – both parts are running the same JavaScript code and use the same service as a source of data. For this purpose I created Catberry.js – a framework for creating isomorphic apps using node.js which was eventually open-sourced on GitHub and got some attention from the JavaScript community. One feature to stand out was progressive rendering versus a common buffered rendering approach.

Novosibirsk, Russia

Working as a .NET‐developer building a high-load service for processing and modifying streams of data. The service was a part of the integration bus system between dozens of internal services/products.

Barnaul, Altai Krai, Russia

Worked in a team building custom contract-based projects (.NET, node.js).Among them:* Push‐notification service that notifies social network users about the beginning of a tournament (node.js, Socket.io).* Administration console for a distributed gambling system (bet on sport events, etc). It allowed users to view detailed analytics about the bets and to control the system remotely (ASP.net).* Web‐application for the text‐translation workflow. It was used as a tool for managing, assigning and solving tasks among professional translators (Ext JS 4).

Novosibirsk, Russia

My assignment was to build a plugin for Microsoft Visual Studio 2010 that would visualise results produced by Intel Static Security Analyser. I proposed a concept of the GUI for the tool, created a UI component for iteration‐based loading and parsing of the large amount of XML data (LINQ‐To‐XML) and built the plugin itself using WPF.

Barnaul, Altai Krai, Russia

Worked as a lead developer of a WebSCADA system that made it possible to remotely monitor and control water pumps and heating systems (any type of a system in theory).The system worked as a subscription‐based web‐app. The subscription included necessary devices and sensors for the customer.I was responsible for the initial architecture and communication protocols and built:* back-end service for real‐time data visualization (node.js, Socket.io).* data visualisation dashboard UI.* back-end service for the management console that calculates analytics and generates reports (PHP).* management console UI (ExtJS).Later I used the results of this work as my diploma thesis.

Novosibirsk, Russia

My assignment was to research the influence of a managed runtime environment (MRTE) on the Intel Math Kernel Library performance using JVM and CLR (.NET) as examples.I've created a set of performance benchmarks for FFT‐related functions from MKL and proposed a validation algorithm for MKL FFT that allows us to simply compare library‐produced results with the expected correct answer of the transformation.