Suphakrit Lertkitcharoenvong Email & Phone Number

United States

Charleston, South Carolina, United States

• Research and perform PoC to compare between Azure Synapse Analytics and Microsoft Fabric
• Re-architect the current system by consolidating data pipelines from Azure Functions, Data Factory, and Databricks to Azure Synapse Analytics Workspace
• Establish Azure Virtual Network, Private Endpoint, Firewall, and RBAC for internal resources to enhance security
• Constructed an event-based change-data-capture Azure Data Factory data pipeline to extract the unstructured data from MongoDB, transform into structured format, and load it to Snowflake

Ithaca, New York, United States

• Project: Detecting Semantic Change Over Time
• Collaborated with a research team under Professor Lilian Lee and Khonzoda Umarova to investigate the semantic shift of words, such as how the word "positive" acquired a negative connotation post-COVID. The research.
• Applied contextual word embeddings from BERT models to analyze semantic shifts across multiple datasets, leveraging cosine distance as a measure of meaning change between time periods.
• Conducted permutation-based statistical tests and implemented the False Discovery Rate (FDR) procedure to ensure statistically significant detection of shifts across vocabulary terms.
• Evaluated model performance using precision@K and Spearman correlation, improving detection accuracy on the real-world datasets on the Liverpool FC subreddit.

New York, New York, United States

• Creates data reconciliation tools (recon tools) to compare and verify data from source Hive and target Redshift via dbt Jinja macros and utilizes dbt Elementary to generate test case report and Tableau to generate the.
• Develops GitHub action workflow to automate dbt deployment and send out email alerts from failed test cases from the recon tools
• Develops a Source-to-Target Mapping (stm) Python-based automation tool to streamline the conversion of business user’s specifications sheet into SQL DDL scripts, significantly reducing manual conversion work

Ithaca, New York, United States

• Led the team to implement a new Python library called “trinity” containing the data cleaning and feature engineering pipelines and an XGBoost model to estimate the treating specialties of Non-Physician Providers (NPP).
• Collaborated with clients to understand their business objectives and creates a detailed analysis report to serve their data needs

Thailand

• Developed and maintained ETL Azure ADF pipelines, SAP SLT pipelines connecting between SAP ECC and SAP HANA, Qlik Replicate CDC pipelines connecting between SAP HANA and Snowflake, and Snowflake stored procedures to.
• Leveraged Azure ADF and Function App to implement automated data validation solution, reducing data discrepancies between SAP ECC and Snowflake platform
• Coordinated with data engineering teams from Houston (USA) and Curitiba (Brazil) to migrate a legacy on-premises data from SQL Server to Snowflake and move legacy gits from Azure DevOps TFS to GitHub
• Created a robust anomaly detection model from Pandas, Scikit-Learn, and SciPy using the time-series Holt-Winter decomposition method to detect abnormalities in the financial reporting, eliminating all manual.
• Led the Analytic Engineering Track in the Data Analytics Community to demonstrate how to transform data, ensure data quality, and construct schema object deployment and data lineage using Data Build Tool (dbt) and.
• Won fourth place in a global internal data analytics case competition with the classification solution involving NLP methods, such as SpaCy, word vectorization, and TF-IDF, and logistic regression model to categorize.

Thailand

• Project: Retail Credit Risk Prediction Modelling Using the Alternative Factors
• Data Wrangling: Harvested data from Impala through SQL and applied text mining methods, such as RegEx, N-gram, Tokenization, Pythainlp, and FuzzyWuzzy, to clean and map the data about companies that loaners work for to.
• Machine Learning and Data Analysis: Developed a random forest model from 30 engineered alternative features via Pandas, Scikit-Learn, and Tableau to predict the NPL loaners in TTB Credit cards, Flash Card, Cash2Go.
• Optimization and Implementation: Hypertuned the random forest model to reach 70% AUROC score, applied the model to further explore the opportunities in 9 million TTB retail product users, and deployed the model in TTB.

Thailand

• Cooperated and assisted with the professors to grade students' homework, quizzes, and exams on 2183213 Mechanics of Material for AI and Robotic Engineering

Thailand

• Taught two classes of 10th grade and 11th grade with 40 students in each class in high school Physics subject on topics including: 1. Mechanics2. Oscillations and Mechanical Waves3. Electricity and Magnetism4. Light.

Bangkok, Bangkok City, Thailand

• Project: Vehicle Production Volume Prediction and Efficiency Prediction
• Data Analysis: Developed a real-time model via SQL and Python to predict the efficiency and number of cars produced in a single day from absenteeism, machine downtime, line balance, and vehicle models data from.
• Data Visualization and Business Implementation: Established and implemented an effective Tableau dashboard visualization into the BMW Tableau server to be presented on the television real-time dashboard inside the.
• Successfully completed this 6-month planned intern project within a 2-month time schedule with the vehicle volume prediction model attaining 88% accuracy within +/- 1 car

Bangkok, Bangkok City, Thailand

• Project: Data Analysis of "LDA: Ladies of Digital Age" Facebook Page (1.2M followers) Backend Data
• Data Analysis: Executed several data analytics projects requested by the marketing team, including "Most appropriate time for posting on LDA Facebook page", "Relationship between KOLs (Key Opinion Leaders) and post.
• Data Visualization and Business Implementation: Oversaw, analyzed, and created a visualization report of the backend data of the "LDA: Ladies of Digital Age" Facebook Page (1.2M Like) to uncover the insight into social.

Master'S Degree, Applied Statistics And Data Science, 3.92/4.00

Activities and Societies: Cornell Thai Association• Granted Full Bank Scholarship of over 4 million Baht (Declined the offer) • Received.

Bachelor Of Engineering - Be, Automotive Engineering Technology/Technician, 3.89/4.00 (First-Class Honors With Gold Medal Award)

• International School of Engineering (ISE) student • Major: Automotive Design and Manufacturing Engineering • Full scholarship awarded.

High School Diploma, Science-Mathematics Program, 3.89/4.00

• Yothinburana English Program Student studying Math-Science Program GPA: 3.89/4.00 • Remained in the gifted class (king room) for three.