Explore live interactive architecture drawers, verified certifications, and explainable AI pipelines:
svkhun.vercel.app

"I am a Computer Engineering student at Srinakharinwirot University specializing in Data Science, Machine Learning, and Data Engineering. I focus on developing production-grade pipelines, sub-80ms real-time Graph Neural Networks (RGCN), credit risk assessment systems, and transparent, explainable AI architectures tailored for the financial services industry."

• Education: B.Eng. in Computer Engineering, Faculty of Engineering, Srinakharinwirot University (SWU) (2022 – Present).
• Target Roles: Aspiring Data Scientist / Machine Learning Engineer seeking technical opportunities and internships within Banking, FinTech, and Risk Analytics.
• Core Domains: Relational Graph Intelligence (RGCN / PyG), High-Throughput Sub-80ms Inference (ONNX Runtime), Credit Risk Modeling (PD, Long Overdue Debtor forecasting), Explainable AI (SHAP, LIME), and Production Data Engineering.
• Problem Solving: Multi-competition hackathon finalist experienced in optimizing tabular & graph prediction models under strict evaluation metrics (AUC) and production latency SLAs (P99 < 12ms).
• Languages: Thai (Native), English (Professional Working Proficiency, CEFR B1+ / Oxford Test Score: 51), and Japanese (Beginner).

• KBTG Kampus Hackathon 2026 -- Track 2: Data Science & Intelligence (Sep 2026)
  • Architected K-Sentinel & WealthPilot, an industrial-grade K PLUS digital banking copilot combining autonomous cashflow optimization with sub-80ms real-time relational graph intelligence.
  • Built a 2-hop Relational Graph Convolutional Network (RGCN via PyG) to detect multi-tier mule account rings and task scams; optimized model via ONNX Runtime to achieve P99 inference latency of 11.62ms (exceeding the strict <80ms SLA).
  • Engineered autonomous Safe-to-Spend forecasting, automated micro-sweep savings vault (2.0% p.a.), and deployed a full-stack production architecture with FastAPI and React 18 Router SPA on Render.
• Aihack Thailand 2025 (Final Round) -- Organized by AIRA & AIFUL, Chulalongkorn Business School (Dec 2025)
  • Advanced through the Selection Round to compete as a national top 8 finalist in an intensive 3-day data science competition.
  • Developed machine learning classification models on real-world financial loan datasets to predict Long Overdue Debtors (LOD, >90 days overdue after 12 months).
  • Engineered predictive features, optimized pipeline performance for AUC leaderboard rankings across Public and Private test datasets on Probspace, and delivered strategic business insights.
• True Innovation Launchpad 2026 -- AI & Data Science Lead (Team Nakphatthana Tuapralat) (Jan 2026)
  • Spearheaded data strategy and predictive modeling for elderly frailty prevention in the Healthcare Innovation Track, developing sub-30ms clinical risk scoring APIs.
• Geospatial Intelligence for Resilience Hackathon 2026 -- Spatial Data Scientist & ML Modeler (Jan 2026)
  • Organized by GISTDA x KMITL x KMUTT; built "Rain-to-Flood" spatial model fusing Sentinel-1 SAR soil backscatter, Copernicus GLO-30 DEM, and precipitation APIs for 30m flash flood mapping.
• CDG Hackathon 2026 -- Team GrandGuardianAI (Jul 2026)
  • Competed in architecting specialized AI solutions targeting high-impact public sector and operational challenges.
• LINE MAN Wongnai Junior Case Competition -- Strategic Growth & Unit Economics Lead (Team Low Cortisol) (Nov 2024)
  • Formulated campus delivery optimization and "Silent Mission" growth campaign with 8.87 THB/user CAC, acquiring 462 VIP trial conversions.