Azizbek Siddikov

• ▸Developed a computer vision object detection model using YOLOv11 on 100K+ images, contributing to 2nd place among 650 companies in AI Champion 2025 (800K USD Achievement).
• ▸Refactored order page architecture and frontend logic, reducing code volume by 4× and simplifying new product line rollouts
• ▸Migrated payment system from periodic polling to a webhook-based architecture in the CRM, improving Stripe integration reliability and system cleanliness.
• ▸Built Python-based automation scripts for IoT device calibration and PCB validation, reducing manual test steps from 6 to 2 and enabling multi-device parallel QA.

• ▸Built and launched a social platform for book discussions ('Twitter for books') from the ground up with a 3-person team, reaching 70+ early users.
• ▸Integrated AI-driven recommendation logic that suggested books based on friends' reading activity and user searches, enhancing engagement potential.
• ▸Improved development reliability and prototyping speed by modularizing core backend components and actively using AI tools throughout development.

• ▸Optimized large-scale image and video dataset pipelines for AI model development, improving labeling accuracy and team coordination through automated progress dashboards.
• ▸Built an internal Python library to automate data preprocessing and file management, cutting dataset preparation time by 60% and improving data consistency across projects.
• ▸Led a team of three data labelers, establishing standardized annotation protocols and ensuring high-quality results across CVAT and MakeSenseAI platforms.

Full-stack platform for creating, managing and joining community-based events.

Full-stack marketplace for clothes.