Teodor

TEODOR POSHYVAK
Python Developer
[відкрити контакти](див. вище в блоці «контактна інформація») | [відкрити контакти](див. вище в блоці «контактна інформація») | Ukraine

PROFILE

Python Developer with hands-on backend and computer vision experience, building production systems end-to-end —
from Django/REST services and databases to real-time video pipelines and model deployment. Delivered object detection,
multi-object tracking, face recognition and vehicle-analytics features integrated into backend services, alongside
Java/Spring Boot REST APIs for e-commerce platforms. Comfortable owning the full stack: backend architecture,
RTSP/WebRTC video ingestion, GPU-optimized inference, and shipping features that reach real users. Looking to bring that
end-to-end delivery experience to a team building applied software products at scale.

WORK EXPERIENCE

Backend & Computer Vision Engineer — Smart Techno Lab, Ukraine
May 2025 – Present
• Designed and shipped computer vision solutions for intelligent video-surveillance systems deployed in production, not
just proof-of-concept.
• Built real-time object detection and multi-object tracking pipelines handling live RTSP camera feeds end-to-end.
• Delivered vehicle-analytics capabilities — including vehicle classification, brand recognition and License Plate
Recognition (LPR) — used for transportation and traffic monitoring use cases.
• Implemented face-recognition and watch-list monitoring for public-safety applications.
• Developed computer vision modules for traffic analysis, queue monitoring, people counting and event detection,
turning raw video into actionable alerts.
• Integrated computer vision models directly into production backend services and optimized inference for GPU-based
environments, improving throughput and reliability.
• Built scalable computer vision services in Python/Django, with PostgreSQL, RabbitMQ, WebRTC and FFmpeg powering
the real-time video pipeline.
• Owned the full model lifecycle: dataset preparation and annotation, training, evaluation and deployment of
production-ready models — then debugged and tuned pipelines for stability under real-world load.

Java Backend Developer — Verocasa, Spain
Apr 2024 – Sept 2024
• Built and maintained backend services with Java/Spring Boot for e-commerce and cosmetic-retail web platforms used
by real customers.
• Designed and implemented REST APIs and backend business logic across multiple live online stores.
• Worked across PostgreSQL and MongoDB to support product, order and customer data.
• Partnered closely with frontend developers to ship complete, working features rather than isolated backend
components.
• Fixed bugs, optimized performance and added new features throughout the software lifecycle — from initial build
through ongoing maintenance.
TECHNICAL SKILLS

Computer Vision: Object Detection, Multi-Object Tracking, Face Recognition, Image Classification, Video Analytics, Vehicle
Analytics, License Plate Recognition (LPR), Model Training & Fine-Tuning, Dataset Preparation & Annotation, Model Optimization
Frameworks & Libraries: PyTorch, TorchVision, OpenCV, Ultralytics YOLO
Video Processing: RTSP, WebRTC, FFmpeg, coturn, Real-Time Video Streaming
Backend & Data: Python, Java, Django, Spring Boot, REST API, RabbitMQ, SQL, PostgreSQL, NoSQL, MongoDB
Languages: Ukrainian (Native), English (Intermediate — technical communication)