Ati Tesakulsiri

Machine Learning Developer | AI Center Unit, Asian Institute of Technology

Skills

• Language: Python
• Tools / Frameworks: Claude Code, Codex, PyTorch, Ultralytics, Hugging Face, Triton Inference Server, Docker, GitLab CI/CD
• Theory: Machine Learning, Deep Learning, Image and Video Processing, Signal Processing, Natural Language Processing, Data Science
• Environment / Operating System: Ubuntu

Working Experience

AI Center Unit

• Machine Learning Developer - Full time, current position
• Software Developer - Full time, after graduation
• Graduate Research Assistant - Internship

Selected Work Projects

• Text-Based Video Smart SearchDeveloping a PoC system using generative AI for Thai-language surveillance-record search. Leveraged LangChain, SmolAgent, and Ollama for on-premise LLM serving; developed vision-language multimodal systems, fine-tuned models, and vector database integration for video analytics programs.
• CCTV Black Smoke Measurement and EvaluationDeveloped research software to detect and analyze black smoke from vehicles in collaboration with ENTEC, NSTDA. Built a web application with video analytics for smoke screening and multi-frame emissions analysis from light-duty diesel trucks and vans.
• Safety AI PoCDeveloping a tollway surveillance PoC for wrong-way driving, speeding, illegal parking, and unexpected object detection. Responsible for a zoom-assisted API for falling-object detection up to 300 meters and vehicle length classification.
• AI Service for Video AnalyticsContributed to AI service tasks, AI training, optimization, hosting tests, CI pipeline upgrades, automated unit and evaluation tests, reporting, and service software patching.
• Face RecognitionOwned the analytics-side face recognition system, including detection, tracking, recognition integration, evaluation, and deployment support.
• Wildfire Simulation and Component ClassificationDeveloped wildfire component classification and segmentation from drone video, with forecasting support using empirical models and satellite data.
• Video Action RecognitionImproved action recognition post-processing to reduce fall detection false positives and performed system performance evaluation.
• Other Demo ProjectsDeveloped accident detection demo logic using tracked object-pair information and Qwen 3.5 for reasoning over vehicle interactions. Built VLM-based scene detection for fire, violence, fall, smoke, and abnormal surveillance events.

University and Personal Projects

• Joules of SaimDesigned and built an end-to-end business intelligence project to forecast electricity prices. Responsible for data pipeline work including data gathering from BOT and EGAT APIs, time-series model training, and deployment of an interactive Python-based web dashboard.
• Diabetes Classification Using Raman SpectrumsApplied traditional machine learning such as SVM and ANN with signal processing techniques to classify diabetes from Raman spectrums.
• Bros of RainbowConducted a comparative analysis of reinforcement learning architectures including Rainbow, DDQN, LSTM, and Transformer variants for OpenAI Gym game-playing agents.
• NLP ApplicationDeveloped and hosted a web application showcasing NLP models, including a resume parser using NER, sequence-to-sequence machine translation, and text generation.
• Serotonin Extraction and Quantitative AnalysisStudied substance absorption with a monolith material approach for urine extraction to perform quantitative analysis of serotonin concentration related to depression.

Thesis

MULTI-MODAL PERSON RETRIEVAL: BRIDGING TEXT, IMAGES, AND RE-IDENTIFICATION

Developing a Transformer-based multimodal system integrating a text encoder and an image encoder for person-related tasks, including identity information processing, cross-camera person re-identification, and text-based person search in one model, following the OpenAI CLIP design.

Education

Master of Science in Data Science and Artificial Intelligence, GPA 3.7/4.0. Recipient of the Hisamatsu Prize.

Courses taken: Recent Trends in Machine Learning, Data Modeling and Management, Machine Learning, Business Intelligence and Analytics, Computer Programming for Data Science and Artificial Intelligence, Natural Language Understanding.

Bachelor of Science, GPA 3.60/4.0. YSTP research scholar from NSTDA and full scholarship recipient from the university.

Scholarships and Awards

• Hisamatsu Prize: Ranked #1 in batch academic record, granted by Asian Institute of Technology.
• Young Scientist and Technologist Program: Granted by National Science and Technology Development Agency.
• Royal Thai Government Fellowships: 100% scholarship granted by Asian Institute of Technology.
• Talented Individuals in Science and Technology Scholarship: Full scholarship for bachelor's degree, granted by Suranaree University of Technology.

Certifications

• CS50x: Core programming and web application concepts with C, Python, and SQLite before master's degree study. Certificate link.
• CS50AI: Core concepts of artificial intelligence and machine learning for master's degree study. Certificate link.