Selected work
Enterprise GenAI Platform
Designed and delivered an enterprise AI platform allowing researchers and clients to interact with proprietary knowledge through conversational AI.
Converted a research prototype into a production application in under two months, enabling researchers to generate product ideas and extract insights without manual document analysis.
Award-winning Agentic AI — Best Paper Award
Designed an autonomous AI system combining company knowledge, client datasets, web retrieval, and domain reasoning.
Best Paper Award — International Market Research Conference, Philadelphia, USA.
Breast Cancer Detection AI
Technical lead for production-grade AI models covering breast abnormalities, density estimation, localization, and clinical deployment.
Improved data quality and deployed clinical-grade inference services with rigorous validation and explainability.
Smart Video Analytics
Designed intelligent video analytics systems for face recognition, human tracking, activity recognition, age/gender prediction, OCR, and digital signage analytics.
Deployed across cloud and edge devices for real-world surveillance and analytics workloads.
AI Infrastructure & MLOps
Architected production ML pipelines with experiment tracking, dataset versioning, automated preprocessing, and CI/CD.
3× reduction in preprocessing time for large-scale cell imaging datasets through pipeline optimization.
Approach
Research to product, not research in a slide deck
I translate prototypes into systems with clear ownership, deployment paths, and user workflows — often under tight timelines.
Architecture before model selection
Retrieval design, data contracts, evaluation harnesses, and operability determine whether AI survives production — not the latest model name.
Lead teams through complexity
Technical strategy, architecture reviews, AI roadmaps, mentorship, and stakeholder alignment across research, engineering, and business.
Measure impact in the real world
Clinical validation, conference recognition, preprocessing efficiency, and products used by researchers and clients — not vanity metrics.
Generative AI. Enterprise RAG, AI Agents, Multi-agent workflows, Knowledge assistants, LLM evaluation, Vector databases, Prompt engineering.
Computer Vision. Object Detection, Face Recognition, Human Tracking, OCR, Activity Recognition, Video Analytics, Edge AI.
Medical AI. Mammography, Chest X-ray, Breast Density, Cancer Detection, Clinical Validation, Explainable AI.
AI Platforms. Azure, AWS, Docker, Kubernetes, FastAPI, MLflow, ClearML, Weights & Biases.
Stack
AI. PyTorch, TensorFlow, LangChain, LangGraph, Transformers, OpenCV, YOLO, ONNX Runtime.
Cloud. Azure, AWS, Docker, Kubernetes, FastAPI, REST APIs.
MLOps. MLflow, ClearML, Weights & Biases, GitHub Actions, CI/CD.
Databases. PostgreSQL, MongoDB, Oracle, SQL Server, MySQL, Teradata.
Languages. Python, SQL, Bash, Ruby on Rails, C/C++.
Writing & talks
- Video analytic system for activity profiling, fall detection, and unstable motion detection
- External Validation of Deep Learning Algorithm for Tuberculosis Detection in Thai Population
- Robust real-time video face recognition system for unconstrained environments
- Automatic Elderly Fall and Unstable Movement Detection System Using Framewise and LSTM Based Video Analytics on an Embedded Device
Additional IEEE conference papers and Perceptra white papers on medical imaging AI. ORCID profile.
- International conference presenter
- Guest speaker on Deep Learning
- National Workshop Mentor
- Graduate Teaching Assistant
- Technical mentor for AI engineers