Seymur Kafkas Email & Phone Number

Milan, It

Milan, Lombardy, Italy

▪ Working on training, productionizing, and serving AI models for Remini, DawnAI, Splice apps. Building and maintaining a model-serving infrastructure/codebase for 100+ million users.▪ Developing AI Backend systems and internal tooling for engineers/researchers (training and inference).▪ Built a self-hosted CI system for the ML Platform, achieved up to 5X speedup in builds, releases, and tests.

Remote

▪ Worked with Python, Express, Flask, Spacy, SQLite, OpenAI API, Beautiful Soup and Hugging Face transformers▪ Developed a resume-parsing microservice. Used a state-of-the-art language model to extract the details into a JSON as per a pre-defined schema. Set up service-to-service authentication, communication and deployed the parser on Cloud Run. Achieved almost human-level accuracy (exceeding 97%); the parser is deployed to production▪ Created an application form filler (Microservice and Chrome Extension). Designed a backend that parses the HTML and sends applicant-specific autofill commands to the Chrome extension. Used procedural, data-driven and NLP methods to answer form questions. The service is deployed on Cloud Run, and has completed 1000+ applications

Istanbul, Turkey

▪ Worked on control of the Baxter Robot and the Allegro Hand. Used CoppeliaSim and Drake for contact simulation, model-based optimization and system design. Used PyTorch extensively as the deep learning framework.▪ Experimented with state of the art Reinforcement Learning algorithms at present to derive robust, model-free control policies.

Istanbul, Turkey

▪ Used C++, C#, and technologies such as SQLite and Oracle DB. Developed high performance data converters for proprietary raster file formats. Used multi-threaded work-queues for parallelizable, compute-bound conversion logic.▪ The converters achieved up to 10X speedup relative to previous implementations, and are deployed to production on 137 servers▪ Set up a monitoring system using Zabbix across a network of 60+ CentOS servers. The central server gathered metrics such as CPU usage, disk usage, and log files from all managed servers; used these metrics in declaring rule-based triggers that reported anomalies in the system as events.

Bachelor Of Science - Bs, Computer Engineering, 3.70 / 4

Courses: Artificial Intelligence, Robotics, Database Systems, Parallel and Distributed Systems, Software Engineering, Computer.

Bachelor Of Science - Bs, Electrical Engineering, 3.73 / 4

Courses: Signals and Systems, Control Systems, Electronics, Circuit and System Analysis, Digital Circuits, Engineering Mathematics.

International Baccalaureate (Ib) Diploma Programme, 38/45

▪ Overall grade of 38/45 with 666 at Higher Level ▪ Subjects studied include Mathematics, Physics, Economics, English A: Language &.