Ievgen G.

Design and deploy AI pipelines that transform image, video, and audio data into reliable real-world applications, from computer vision analytics and 3D perception to audio (speech, music, sound) analysis and generative AI systems. My work focuses on production AI systems, not research prototypes. I help companies move from early feasibility studies and PoC development to scalable deployments running in real environments. Typical projects include: • computer vision pipelines for detection, tracking, and segmentation • real-time video analytics and edge AI deployments • 3D vision and spatial perception systems • speech processing and audio AI solutions • generative AI pipelines for image, video, and audio Computer Vision Systems Design and development of advanced computer vision pipelines for image and video analysis. Typical solutions include: • object detection and multi-object tracking • semantic and instance segmentation • pose estimation and motion analysis • OCR and document understanding • visual search and recognition systems These systems are used in industries such as manufacturing, sports analytics, retail, healthcare, and security. 3D Vision & Spatial AI Development of AI systems that understand spatial structure and depth. Experience includes: • structure-from-motion (SfM) • photogrammetry pipelines • depth estimation models • NeRF and neural rendering • point cloud processing and 3D reconstruction Applications include robotics, AR/VR, construction analytics, and digital twins. Edge AI & On-Device ML I specialize in deploying ML models on mobile and embedded devices where latency, memory, and power constraints are critical. Typical optimization techniques include: • model quantization and pruning • architecture optimization • real-time inference pipelines • deployment on mobile and embedded hardware Technologies include: TensorRT, TensorFlow Lite, CoreML, ONNX Runtime. Many deployed systems operate with 50–100 ms inference latency depending on hardware. Generative AI for Vision & Video Development of generative pipelines for media processing and synthetic data generation. Typical solutions include: • image and video generation pipelines • diffusion-based editing and enhancement • synthetic dataset generation for model training These tools help accelerate AI training and improve model robustness. Audio & Speech AI Development of AI systems for speech processing, audio analysis, and voice technologies. Examples include: • phoneme segmentation and pronunciation analysis • speech recognition pipelines • voice feature extraction and audio analytics • generative audio and music models These systems are used in: • language learning platforms • speech therapy tools • voice biometrics systems • music AI applications Technical Stack Frameworks & Models PyTorch, TensorFlow, OpenCV, Detectron2, MediaPipe, YOLO, DINO, SAM, CLIP Deployment TensorRT, TensorFlow Lite, CoreML, Docker, ONNX Runtime, FastAPI Programming Python, C, C++ 3D Vision NeRF, SLAM, Dust3r, point clouds Leadership & R&D I lead an R&D-focused AI team at It-Jim, an AI consulting company with 30+ engineers and 10+ PhDs specializing in: • Computer Vision • Generative AI • Audio & Speech AI • Edge AI systems We help companies solve technically challenging AI problems and build reliable production systems. If you are looking for experienced AI engineers to design, prototype, or deploy advanced machine learning solutions: feel free to reach out!