Hiep Tran

AI Engineer

Torilab

📍 Hanoi, Vietnam

Developing and maintaining multimodal AI companion with capabilities for chat conversations, image sharing, and visual understanding through camera interactions

Key Responsibilities:

• •Conducted research on controlling Large Language Model behavior without explicit prompting through activation steering techniques
• •Developed and maintained multimodal AI companion integrating vision-language models for understanding images and natural conversations about visual content. Implemented real-time camera interaction system enabling users to share images and receive intelligent responses based on visual understanding
• •Architected end-to-end RAG pipeline integrating document ingestion, chunking, embedding, vector storage, and LLM-based answer generation for financial reports and market data
• •Built chat interface with multimodal context awareness, supporting seamless transitions between text and image-based conversations
• •Performed Large Language Model and vision model fine-tuning including SFT, DPO, and multimodal alignment tasks, with scalable model serving using vLLM and SGlang
• •Developed core framework supporting long-term memory, tool usage, dynamic prompts, and multi-turn multimodal interactions
• •Worked on additional AI capabilities including user emotion detection from text and images, image generation with diffusion models and LoRA, and multi-speaker diarization

AI Engineer

VCCorp Corporation

📍 Hanoi, Vietnam

Developing and maintaining AI systems for Vietnamese document understanding and computer vision applications for social media platforms.

Key Responsibilities:

• •Developing and maintaining Automated Information Extraction module for Vietnamese document understanding, often achieve 99% of accuracy
• •Developing and optimizing automatic face attendance system with face detection and recognition, mask wearing check, person tracking modules
• •Development of image and video classification system for Lotus social network

AI Engineer

A.L.I.S VIETNAM

📍 Hanoi, Vietnam

Developing a system to convert speech and text into sign language to assist Deaf people in accessing information.

Key Responsibilities:

• •Study of speech-to-text, text-to-speech models, GANs, recognition and conversion models to human actions
• •Implement pipeline for machine learning model using FastAPI, Docker, Kubeflow
• •Research and development of multimodal AI systems for accessibility
• •Design and implement real-time processing systems for sign language conversion

Viblo Algorithm Content Creator

Sun* Vietnam

📍 Hanoi, Vietnam

Contributing articles and algorithmic challenges on the Viblo ecosystem, focusing on AI/ML education and research.

Key Responsibilities:

• •Contributing articles, algorithmic challenges on the Viblo ecosystem
• •Research and publish articles on Computer Vision, Natural Language Processing, Recommendation system and MLOps
• •Creating educational content for the developer community
• •Sharing insights on latest AI/ML trends and technologies