Professional Summary
Wassim Swaileh, PhD
AI Engineer -- Robotics, Autonomous Systems & Multimodal AI
About Me
AI Engineer with 8+ years post-PhD experience building production AI models and training pipelines for multimodal and autonomous systems. Expertise in VLMs, vision-language-action models, 3D scene reconstruction, and large-scale ML infrastructure.
Currently working at Huawei Suomi Finland Research Center as a Senior Researcher in Multimodal AI, where I:
- Build and ship production-scale video generation models using VLMs and diffusion architectures
- Design end-to-end training pipelines with containerized environments on ModelArts MLOps platform
- Co-developed ReWind (CVPR 2025), a large language model for long video understanding
- Spearhead OCR capabilities, deploying multimodal perception models across multiple languages
My research spans:
- Video Understanding & Generation
- Vision-Language Models (LLaVA, CLIP, QwenVL)
- Vision-Language-Action (VLA) Models
- Generative Diffusion Models
- 4D LiDAR Generation & Autonomous Systems
- 3D Scene Reconstruction
- Kinematic Time Series Analysis
- Document AI & OCR
Core Expertise
Video AI
Video understanding, generation, and long video analysis with memory models
Multimodal AI
VLMs, VLA models, and multimodal perception systems
Autonomous Systems
4D LiDAR generation, sensor fusion, and 3D reconstruction
Career Highlights
Building production-scale video generation models and co-developing ReWind (CVPR 2025) for long video understanding.
Deep learning pipelines for kinematic time series analysis and trajectory reconstruction using TCNs.
UNet-based architectures for visual perception and 3D modeling of ancient maps.
Production DNN pipelines for named entity extraction from historical documents (EURHISFIRM project).
Key Achievements
CVPR 2025 Publication
Tier-1 ConferenceCo-developed ReWind, a large language model for long video understanding with instructed learnable memory.
Patent Submitted
AI-JournalistApparatus to Generate Media Contents Conditioned to User Preferences Settings.
Get In Touch
Interested in collaborating or learning more about my work?
Contact Me