Yakhyokhuja Valikhujaev
AI/ML Software Engineer
Exploring Life Through the Lens of a Computer Scientist.
Experience
Present
PhD in Computer Science and Engineering
Sungkyunkwan University
education
PhD in Computer Science and Engineering
Sungkyunkwan University
Doctoral research at the Information Research Laboratory (InfoLab), College of Computing and Informatics, advised by Prof. Tamer Abuhmed.
Focus: Facial privacy protection, adversarial machine learning, and identity protection against face recognition and generative AI misuse.
MLOps Engineer
Thaki Cloud Co. Ltd
work
MLOps Engineer
Thaki Cloud Co. Ltd
- Designing and maintaining Kubernetes clusters for private cloud platform supporting distributed model training and inference.
- Building systems for resource discovery, policy enforcement, and workload scheduling using Kubeflow, Kueue, Kai Scheduling, and Kyverno.
- Managing ML pipelines with Helm and ArgoCD for continuous delivery across staging and production environments.
AI/ML Research Engineer
Aria Studios Co. Ltd
work
AI/ML Research Engineer
Aria Studios Co. Ltd
- Fine-tuned Qwen-2.5 models (7B and 3B) for Korean virtual characters using LoRA and DPO.
- Developed real-time voice-to-voice assistants with LiveKit and Whisper, integrating MCP servers for task execution and Phi-3-Vision for visual understanding.
- Added semantic search over virtual characters using embeddings indexed in FAISS and Pinecone.
- Built inference APIs for self-hosted LLMs (vLLM) and Stable Diffusion models, using LangChain to convert user requests into structured inputs for image and music generation.
- Fine-tuned FLUX with LoRA to generate movie characters with a consistent appearance across scenes.
ML Engineer
Pyler Co. Ltd
work
ML Engineer
Pyler Co. Ltd
- Built video and image content moderation pipelines for brand safety.
- Developed real-time object detection and multi-label classification models to identify unsafe content.
- Clustered datasets using CLIP embeddings, PCA, KMeans, and DBSCAN to support data curation and annotation.
- Trained and evaluated models on Kubeflow Pipelines, applying active learning to prioritise samples for annotation.
AI Research Engineer
D-Meta Co. Ltd
work
AI Research Engineer
D-Meta Co. Ltd
- Developed OCR pipeline using Spatial Transformer Networks (STN) achieving 90%+ accuracy for industrial scene text.
- Built ANPR system with real-time RTSP stream inference improving performance by ~15%.
- Deployed car damage detection on Android devices using TorchScript.
- Applied Pix2Pix GAN for shadow removal.
MSc in Computer Engineering
Gachon University
education
MSc in Computer Engineering
Gachon University
Master’s degree in Computer Engineering, advised by Prof. Young Im Cho. CGPA: 4.0/4.5
Graduate Research Assistant at the AI and SC Lab, working on computer vision and lightweight models for edge devices.
BSc in Computer Engineering
Tashkent University of Information Technologies
education
BSc in Computer Engineering
Tashkent University of Information Technologies
Bachelor’s degree in Computer Engineering. CGPA: 85/100 (3.72/4.0)