Computer Engineering Student at Chulalongkorn University (CEDT)
  Architecting the end-to-end synergy between high-performance data engines and autonomous agentic systems.

   

• Solid Foundations First: Prioritizing deep mastery of DSA and OOP. I believe that architecting elite AI systems requires the ability to build core infrastructure components from the ground up.
• Unified Architectural Ownership: I don't just build pipelines; I design the entire ecosystem. This means ensuring that the Data Infra (the circulatory system) and the AI Agents (the brain) are architected as a single, secure, and performant unit.
• Production-Grade Agency: Moving beyond simple prompts by engineering robust agentic guardrails—integrating PII masking, AST-based validation, and semantic memory directly into the platform architecture.

• Languages: Python (Systems), C++ (DSA), Java (OOP), SQL (Analytics)
• Data Engineering & Orchestration: Apache Airflow, Prefect, dbt, Apache Flink (Streaming)
• Data Architecture & Lakehouse: PostgreSQL (Catalog JDBC Backend), Apache Iceberg v2 (Merge-on-Read), Apache Kafka (KRaft), Trino Engine, MinIO S3
• AI Agent Platform Layer: OpenAI API, ChromaDB (Vector Semantic Cache), LangChain/LangGraph, Streamlit, Plotly
• Governance & Quality: Salted SHA-256 PII Masking, Schema Validation, Automated Metadata Tracking

A full-stack architecture transforming traditional e-commerce data into an autonomous BI ecosystem.

• Core Mechanisms: Designed an end-to-end flow where Prefect manages data movement while an AI Analytics Agent performs reasoning over complex star schemas.
• Architectural Focus: Engineered a sub-0.1s ChromaDB semantic cache and an upgraded AST-based SQL validator to ensure the agent's decision-making is both high-speed and secure within DB boundaries.

A high-throughput streaming architecture providing real-time context for financial AI reasoning.

• Core Mechanisms: Architected a PyFlink event-time windowing system over Kafka and Iceberg v2 to provide a consistent "Source of Truth" for downstream agents.
• Governance Focus: Enforced real-time Data Governance via Salted SHA-256 masking at the ingestion layer, ensuring privacy is maintained before data enters the AI's reasoning window.