BCR, Bucharest

During my assignment to BCR I was responsible for developing and leading six to ten in-house projects, as well as implementing change requests on several business-critical legacy applications. Covered FE & BE development, project leadership and system design implementations. I was also holding knowledge sharing sessions regarding Angular usage and Java coding best practices, as well as introducing the team to various automated testing methodologies and implementations. I also helped the team to adopt a company-wide CI/CD lifecycle for applications that were to be deployed on Openshift via Gitlab and Argo.

Some of the projects worth mentioning are:

A hub to serve as a single point of interest for all of the bank's backoffice activities. Used Angular module federation to allow integration with various Angular-based clients representing different bank business areas and websockets to populate the dashboard with real-time task statuses.

A frontoffice-to-backoffice ticket management system containing various clients' requests that need to be processed by different bank business areas. My role was to lead all development areas and also manage the migration from the old BPMN implementation.

A buffer between the bank and the national Credit Bureau. Every credit report enquiry to the Credit Bureau is being saved in a local database to be reused in a specific timeframe, if the same query is being performed for the same person nation-wide, for cost purposes. I've been leading both the frontend and backed development, as well as making sure the consumers (other various internal bank systems) are integrated properly via SOAP or REST.

review.center

review.center is a personal start-up initiative using AI that would later become Software-As-A-Service. It facilitates 360° employee performance tracking with Natural Language Processing for reviewer Sentiment Analysis.

I covered the entire life cycle of the software product from idea to reality. During this time I also had the chance to learn, experiment and build neural networks. I attempted improving neural net design flows by experimenting with custom-build structures and propagation methods. I ended up using Google's API in the end since it has the best and most consistent sentiment accuracy.

The solution is a web based application that focuses on Employee Performance Management. The system allows employers to track performance feedback on projects and/or time-based intervals supporting the concept of 360 feedback. The twist is the AI part: automatic user sentiment is being extracted based on what feedback is given, thus offering the possibility of a completely automated review cycle with no management input required. The solution also tracks and graphs out aggregated employee performance data to support yearly compensation and promotion evaluation.

A flexible permission scheme allows multi-level access to private employee performance and feedback, allowing all levels of an organisation the correct access as employee, manager, and HR admin.

Reminder notifications are being sent out to the relevant people when something needs being addressed in the review flow, so the user can remain focused on more important tasks.

The solution is hosted on Amazon Web Services (AWS) and it is build using Java 1.8 with Spring (Boot, Data, Security, Profiles), Javascript (+jQuery), scheduled tasks, REST services and relational MySQL and/or H2 in-memory persistence.

PROX

Mobile app for gamers.

I was responsible for all aspects of backend development and deployment, using AWS, Java, Spring and some passion.

You can download it here for both Android and iOS!

ING, Amsterdam

I was part of a team responsible with designing a strategy to reduce and completely eliminate the hassle of API mismatch between system-wide applications, thus drastically improving the delivery time as well as the QA factor.

Building on top of PACT consumer-driven API contracts concept, I was responsible with building proof of concepts and offering on-site implementation support across various use-cases, for teams inside ING across The Netherlands and Benelux areas.

The tech stack includes Java 8, jUnit, PACT, Spring API-related frameworks and some Scala and Kotlin integrations on the side.

Financial Times, Amsterdam

I have been working alongside around 15 other developers at improving the current codebase of the FD Mediagroep, using the latest technologies and frameworks, as well as implementing new software solutions to serve their needs.

Most of my work was around the biggest content websites of the Mediagroep (fd.nl, bnr.nl and esb.nu), during changing the backbone of the codebase to a more component-oriented structure, as well as offering support for a new CMS.

Made use of cloud computing solutions to achieve a much more reliable, fast and scalable environment, while still ensuring seamless integration with a vast number of external services.

• AWS stack: EC2 linux instances (~40, different sizes, including DTAP and CD), RDS (MySQL), S3 storage; also includes IAM configurations, general access roles and setup.

• Server architecture created and managed in Ansible.

• Java multi-tier backend wrapped in Spring, structured in multiple different-scoped projects and modules, with a single codebase approach.

• Performant and scalable frontend structured on Thymeleaf, with TypeScript and Node.js.

• Continuous delivery environment supported by Gitlab, Jenkins and Rundeck with multi-repository configurations.

HCL Technologies, Amsterdam

I was responsible with leading the development of a java-based assessment framework, similar in concept with Hackerrank or Codility. It can evaluate the Java, Scala, devOps or UI skill-sets of any new candidates that will join their clients. It uses fully-automated CDaaS pipelines for each technology it supports.

I was managing deployment and configuration of all client instances for the above framework. By using AWS I've managed to reduce new instance creation and full configuration time to hours instead of days.

I also hosted technical interview sessions for offshore Java and DevOps resources via phone or face-2-face interviews.

Development of several POCs along the way; worth mentioning is an API made to validate Swagger docs, using custom rule-sets that can be customized by individual clients.

I was also assigned to work as a consultant for one of the main clients (see above).

Dell - Dell Financial Services, Bucharest

I was hired to be part of a roadmap to migrate from core legacy systems to a microservice architecture covering several domains.

The legacy systems were java-based and the microservices were to be created using .NET Core 2. Working with the overseas architecture and integration teams, I've successfully developed and deployed some of the first microservices across the Account domain via Pivotal Cloud solutions and Gitlab pipelines.

In the meantime the team and I were also maintaining and developing new functionality for the core legacy applications. I've also participated in upgrading the webserver version across a couple of applications, which does not sound much, but when it comes to legacy it can become really scary.

The microservices are using custom caching methods and queueing services implemented using distributed Redis clusters and RabbitMQ, as a requirement to deal with external service outages and act as buffers for other core applications across DFS. They were also a step into introducing BDD and TDD techniques across the team, with fully-automated functional and unit testing via pipeline execution and deployment.

Transilvania University, Brașov

Mechatronics, Robotics, and Automation Engineering