Aiven is a cloud agnostic DBaaS, aiming to build a trusted open source data platform for everyone. As a member of the Data Streaming team, I was responsible for the Kafka and Flink offerings, as well as related services and ecosystems.- Improvements in the system tests of the Kafka service(s) focusing on reducing flakiness and better performance- Improvements in Flink application management, speeding up API performance by 10x in certain use cases- Implement OAuth2 and OIDC capabilities in the Kafka offering, and the (open source) Karapace REST proxy service- Replace the various native Python Kafka libraries in the Karapace REST proxy, schema registry, and backup tools with the librdkafka-based confluent-kafka-python - this included implementing AsyncIO wrappers around Kafka clients

Following Shopify's acquisition and introduction of Deliverr into Shopify Logistics my team has kicked off a new project to build a first-party EDI solution to enable an easy, industry-standard way of handling wholesale orders. Electronic Data Interchange (EDI) is the computer-to-computer exchange of business documents in a standard electronic format between business partners.- Technical project kickoff, establishing CI/CD, production infrastructure, used technologies, and automated testing principles (AWS lambdas and step functions, TypeScript, TypeORM, CircleCI)- Integration with a 3rd party EDI translation service

Shopify Fulfillment Network (SFN) is an in-house logistics solution to store merchants' inventory and fulfill their orders without them having to worry about or actively manage their fulfillment. My team in the Growth area of SFN was responsible for the smooth, self-serve onboarding of merchants into Shopify's network of warehouses. This spanned multiple projects as we have successfully worked towards the full automation of bringing new merchants on board. All projects required full-stack development, including a back-office management tool.- Implementation of self-serve contract signing within the SFN app, including data-tracking - this eliminated a lengthy manual process of sending documents back and forth between the merchant and the company (Ruby on Rails, TypeScript, React, and Mode)- Technical design and implementation of a task-based onboarding process, replacing the existing linear one- Championed, designed, and implemented a project for the self-serve setup of shipping rates - a previously manual process that took hours for the onboarding-specialist support team- Championed, designed, and implemented a project for self-serve go-live of merchants with Shopify's SFN offering, making 95% of support contact redundant - now most merchants are able to onboard on their own, using the SFN app without manual intervention from the support team - designed a new manual process for exceptional cases where white-glove services are required, negotiating with various stakeholders, including the support team- Took the lead on the SFN application's on-demand cloud environment setup for development - introduced performance improvements, additional seed data, and configuration options for a smooth developer experience, saving 20 minutes per day per developer on average- Established testing and development principles for Shopify's in-house URL shortening tool, along with observability, allowing for production readiness and easy-to-use URL shortening for other systems

Tech lead of Public Cloud Value Added Services, a team responsible for logging, metrics, alerting, and backups of customers' applications in Mendix's low-code/no-code PaaS product. The Mendix Public Cloud is built on top of AWS (S3, RDS, Kinesis, SQS, etc.) with backend services deployed in multiple Cloud Foundry clusters: the majority of these services are written in Python, with some in Go and Java.- Lead the effort of re-architecting the entire logging stack of the Mendix Public Cloud, from an AWS S3 based solution to using ElasticSearch- Act as a point of contact for team members regarding architectural decisions while actively participating in software development- Establish quality and architectural guidelines, along with processes - some of which got adopted by the wider organization later on; such as standardized CI/CD and configuration management, release approvals and compliance, and code review guidelines

Tech lead of the Licensing and Metering team - a temporary assignment to bootstrap a team exclusively composed of people new to our company. The team would successfully automate previously manual licensing workflows and introduce metering for billing purposes.

Member of the team responsible for the so-called value-added services of the Mendix Public Cloud. This includes logging, backups, metrics, and alerting functionality for customers' low-code/no-code Mendix applications.- Maintenance of existing legacy technology stacks- Implement full CI/CD for the above-mentioned services- Implement full Infrastructure as Code for the above-mentioned services- Integrate IaC with application configuration managementThe CI/CD (Jenkins and GitLab CI) and IaC (Terraform) automation enabled Mendix to launch a new product offering, Mendix Cloud Dedicated, which allows customers to run the entire Mendix Cloud infrastructure in their own AWS account.

As the PO of the mobile app development team, I was managing the work of an external team of iOS and Android developers as well as UI/UX designers, liaising with the internal development team and stakeholders.- Roll-out of improved payout features- Complete redesign of the mobile applications- Introduction of business metrics and A/B testing

Jedlix is a smart charging solution for electric vehicles. Using the Jedlix mobile app users can charge their cars at optimal times when electricity is cheap and energy from renewable sources is in abundance. By doing so, users earn a profit since Jedlix participates in balancing the electricity market.- Integration and acceptance testing of microservices (Javascript, Cucumber JS)- UI testing of iOS and Android mobile applications (JavaScript, Appium, WebdriverIO)- Introduce and integrate a "mock-Tesla" application for testing purposes- Share and introduce testing best practices in the development team (TDD, BDD)

Coaching developers in the field of test automation - providing technical and test design guidance - while making sure we focus on building the right software and building the software right.Implementing unit, integration, acceptance and UI tests and building tools for test automation with Python, Robot Framework, Selenium, Javascript and TestCafe in Jenkins pipelines.The project reached about 80% feature coverage, reducing day-to-day manual testing efforts by the same amount.

As a back-end developer I was responsible for the following functionalities of Bynder Orbit (https://www.bynder.com/en/orbit/): - Credit card payments, subscriptions - end to end, from signup to invoicing - GDPR compliance (right to be forgotten, right of data access/portability, right of rectification) - Authentication and authorization (including single sign-on, account, and user permissions) - Account management - User management - Group management - Portal customizationThe back-end services were mainly written in Python using Pyramid, GraphQL (Graphene), Celery, and SQLAlchemy, and deployed on AWS, making use of containerization (Docker) and Ansible for configuration management. Used PostgreSQL, ElasticSearch and Redis in the persistence layer.

Development in operation and maintenance area of Ericsson's IaaS cloud- Designed and implemented unit and feature tests of legacy upgrade framework, and integrated them in CI (Python, RobotFramework, Jenkins, Groovy)- Extended deployment framework with the virtualization of certain cloud components (Python, Ansible, Bash)- Participated in the design and implementation of a new update framework (Python, Ansible, Bash)

Development of fault management component of Ericsson's IaaS cloud: - Actively participated in the full lifecycle of the software including architectural decisions, implementation and testing - Designed and implemented unit and feature tests (Python, RobotFramework) - Member of the Robot Framework group, providing guidelines and reusable solutions to the development team for system and component testing - Extended REST API microservices of fault management (Python) - Collaborated with continuous integration teams to enhance the building, packaging, and integration of fault management into the cloud system (Python, Jenkins, Debian packaging, Puppet)

RefactorErl is a static analysis tool for Erlang software written in Erlang, developed in a university project at Eötvös Loránd University: - Bachelor's thesis: Implementation of a new GUI for RefactorErl (Erlang, C++, Qt) - Master's thesis: Static analysis of Erlang's generic server behaviour (Erlang)

A demonstrator is a student helping out with and participating in teaching and other related activities of university courses.My duties included:- Developing, handing out and grading of homework assignments- Both private and group consultations for a given course- Teaching seminars - direct practical application of lecture materialI was a demonstrator for the following classes:- Specification and Implementation of Distributed Systems (Bsc)- Synthesis of Distributed Systems (Msc)- Type Theory (Msc)

Developed GUI applications for system testers in Python and PyQt for: - Analysing large log files - Management of test databases - Monitoring of telecommunication devices during test runs - Emulation of telecommunication devices (for the purpose of testing the robustness of certain software components that communicate with said devices)

The Neural Information Processing Group at Eötvös Loránd University focuses on artificial intelligence research and its applications in education and psychology. The group has members from the fields of computer science, physics, mathematics and psychology as well. My duties were the implementation and fine tuning of algorithms and the development of software tools used for image processing and time series analysis, for example to analyse facial expressions, using Python and MatLab.