Kristijan Stepanov

Kristijan Stepanov

About

Detail

Sr. Software Engineer at DeepIntent
Banja Luka, Republika Srpska, Bosnia and Herzegovina

Contact Kristijan regarding: 
work
Full-time jobs

Timeline


work
Job
school
Education
auto_stories
Publication

Résumé


Jobs verified_user 0% verified
  • DeepIntent
    Senior Software Engineer
    DeepIntent
    Feb 2026 - Current (6 months)
    - Backend Development (worked on high-performance real-time bidding system, optimizing code, and utilizing Docker/Kubernetes. Proficient in MySQL, Aerospike, BigQuery, Spark, Kafka, and cloud storage solutions; Deep understanding of Spring Boot for web applications, with expertise in configuration and troubleshooting). - QA Automation (created custom integration testing framework on Docker/Kubernetes and authored unit tests using JUnit) - Analytics (integrated Metaflow for analytics, simplified script maintenance with Python, and migrated scripts from crontab)
  • DeepIntent
    Software Engineer
    DeepIntent
    Jun 2023 - Feb 2026 (2 years 9 months)
    • Designed and implemented a smart throttling feedback controller to maintain system stability under high load, dynamically regulating traffic based on CPU utilization and latency while balancing stability constraints with business-driven request prioritization. • Built and integrated a smart throttling model (XGBoost-based), performing feature engineering, training, evaluation, and tuning, and applying Bloom Filter–based identity prioritization to improve request selection efficiency • Built distributed data pipelines using Spark and HDFS, generating reporting statistics and feedback-loop datasets for system optimization and ML training, including analysis of Aerospike identity graph data • Led end-to-end development of the Sequential Targ
  • DeepIntent
    Junior Software Engineer
    DeepIntent
    Jun 2021 - Jun 2023 (2 years 1 month)
    • Dedicated significant time to understanding programmatic advertising fundamentals and real-time bidding (RTB) system architecture, while building practical experience with Java, Spring Boot, Kafka, MySQL, Aerospike, BigQuery, Docker, Kubernetes, and CI/CD pipelines (Jenkins) • Contributed to the development and expansion of a Testcontainers-based integration testing framework, improving unit and integration test coverage across interconnected services • Worked on smaller backend services within the bidding platform, delivering scoped feature implementations, bug fixes, and incremental improvements under team guidance
  • Bravo Systems doo
    Junior QA Engineer
    Bravo Systems doo
    Jun 2021 - Feb 2022 (9 months)
    • Employed as a QA Engineer through Bravo Systems and assigned to the DeepIntent engineering team • Contributed to test automation and integration testing efforts within a real-time bidding platform • Gradually expanded responsibilities into backend development tasks in collaboration with the DeepIntent team
  • Q Station
    Introduction to the Data Science
    Q Station
    Oct 2020 - Feb 2021 (5 months)
    Completed a comprehensive data science course covering Python programming, Numpy, Pandas for data analysis, data visualization, and the fundamentals of AI and Machine Learning. Applied these skills through hands-on projects, enabling effective data analysis and informed decision-making.
  • Faculty
    Efee system for Computer centre
    Faculty
    Mar 2020 - Jun 2020 (4 months)
    During my professional practice, I worked as a full-stack developer on the Efee system, which is a system for managing student bulletin boards, graduation theses, lecture schedules, and appointment reservations at the Faculty of Electrical Engineering in Banja Luka, Republic of Srpska. The technologies used in the project included Spring, React, Angular, PostgreSQL, and a bit of C#. The system aimed to streamline the process of managing various academic resources, making it more efficient and user-friendly for students, faculty, and administration.
Education verified_user 0% verified
  • F
    Master's degree, Computer Software Engineering
    Faculty of Electrical Engineering Banja Luka
    Jan 2021 - Dec 2024 (4 years)
    Thesis: Applied machine learning techniques for lithium-ion battery health prediction, including feature engineering, model evaluation, and neural network approaches. Published research presented at INFOTEH conference. Coursework & projects included cloud-native systems (Docker, Kubernetes), distributed systems concepts, and scalable microservices architecture.
  • F
    Bachelor's degree, Software Engineering
    Faculty of Electrical Engineering Banja Luka
    Jan 2016 - Dec 2021 (6 years)
    As I progressed through my college education, I gained a comprehensive understanding of computer science and its various applications. I started with the basics in my first year and expanded my knowledge in subsequent years by studying specific areas such as cryptography, computer networks, web development, human-computer interaction, and machine learning. I also gained experience working in a team and using various tools and frameworks. My graduation thesis was focused on web application development using the Django web framework, where I described the framework in detail and created a music web service called "Kikify" using Python. It was an exciting opportunity to apply all that I had learned throughout the years and showcase my abilitie
Publications verified_user 0% verified
  • r
    Application of Machine Learning Techniques for Predicting the State of Health of Lithium-Ion Batteries
    rd International Symposium INFOTEHJAHORINA INFOTEH
    Apr 2024
    This research applies machine learning techniques to predict the state of health (SOH) of lithium-ion batteries, crucial in modern electronics and sustainable energy systems. It compares various machine-learning methods using the NASA Prognostics Center of Excellence dataset, adopting a unique approach of dividing the dataset based on entire batteries rather than individual charging cycles, coupled with meticulous hyperparameter optimization for each model. It explores MLP Neural Networks, CNN, CatBoost, and XGBoost, evaluating their performance based on Mean Squared Error (MSE), R-squared values, grid search and prediction times. This study offers a comparative analysis of machine learning methods in battery health assessment, highlighting