• Built streamlined & scalable ETL pipelines for 400+ calculated features • Developing scalable pipelines using Java, Apache Flink, and Apache Kafka for tracking 160+ vessels in real time • Using distributed Flink applications to process millions of data points in real time with watermarks • Managing Kafka topics efficiently to ensure reliable and ordered message delivery across multiple consumers• Building automated data pipelines to get data from external sources, resulting in 400% increase in efficiency • Developing, testing, and deploying AWS Lambda functions with Secrets Manager and S3 buckets using Python • Authenticating with external databases on Azure using MS Graph for batch requesting • Orchestrating the functions using Apache Airflow• Building web apps using Python, Flask, JavaScript, and Bootstrap to automate business needs • Automating articles’ scraping and summarization while extracting metrics using pre-built offline LLM models• Building scripts to efficiently scrape millions of articles and data points from Google and Google Trends • Drafting analysis graphs using Pandas and Seaborn resulting in an insightful presentation to top management

• Leading the development team’s projects using Agile methodology• Managing a team of 16 people including developers, and designers increasing project efficiency by 80%• Defining & designing target architectures, tech stacks, and resources distribution for clients’ data projects• Conducted pair programming interviews for Jimdo with 60+ Jimdo Backend engineers candidates. The main focus was building APIs on the fly using any technology while applying testing best practices.

• Building data analysis dashboards for clients’ requests using SQL queries with 1B+ data points• Analyzing and visualizing data patterns and trends to hypothesize and understand complex problems• Building Python notebooks for data cleaning and exploratory data analysis using 5,000,000+ data points• Building & comparing Machine Learning models on this data resulting in deploying it to production

• Learned TensorFlow Recommenders in 2 months and designed a personality-based Deep RS• Implemented the recommender system using Python with 300+ features to achieve 90%+ accuracy• Co-authored 2 provisional patents regarding AI systems we designed• Presented the ML product to Verizon and in Collision Toronto which led to a 120% reach increase

Machine Learning Engineer position where I was responsible for designing an innovative recommender system for advertising, analyzing users' data using ML methods, and doing research in the field of ML for future products and services.

• Implemented an automated cybersecurity service in Python for 5,000+ students and staff• Used APIs, Web crawlers, and SQL to automate the sending of certified emails