Skip to experience
techgirl.nyc

Data science · Machine learning · Optimization

Hi, I’m
Oranich.

Three years turning data into decisions people actually make.
Analysis you can trust, models that make it to production, stakeholders who know what happened and why.
Data scientist, now at Columbia for an M.S. in Operations Research.

Oranich Jamkachornkiat
Oranich JamkachornkiatData Scientist · New York CityM.S. Operations Research · Columbia
Expected December 2026

Work experience.

Jun–Aug 2026

MTA · New York City Transit

Department of Subways · New York, NY

Performance Analytics Intern

  • Connected 6 million train movements with 490,000 track-work records to measure how much longer trains take through work zones: 63.7 seconds per segment, about 70% longer than normal. Showed one department accounted for 89 hours of midday delay, and leadership agreed to shift that work overnight.
  • Built automated data pipelines so teams could update the analysis daily and revisit earlier periods without repeating the manual work.

Nov 2023–Dec 2025

SWAT Mobility

Transport technology · Singapore

Data Scientist

  • Optimized routes for 10,000 daily 7-Eleven deliveries, reducing planning time by 70% and operational costs by 8%.
  • Trained a language model to split DHL Turkey’s free-text delivery addresses into street, building, district, and city fields at 98% accuracy, so routing systems could use them.
  • Analyzed more than 2 million GPS records to help Japanese transit operators reduce inefficient off-peak trips by 30%.
  • Automated daily operational monitoring with Slack alerts, saving each team member about four hours of manual investigation a day.

May–Aug 2022

Shopee

E-commerce · Bangkok, Thailand

Data Analytics Intern (Automation)

  • Automated eligibility checks and built scorecards to help teams review merchant requests for pricing exceptions, cutting turnaround from seven days to one.
  • Evaluated tools that read text from images and improved document image quality, reducing the time needed to verify merchant onboarding documents by 85%.

Jun–Nov 2021

Agoda

Travel marketplace · Bangkok, Thailand

Data Analytics Intern (Supply)

  • Built a dashboard to help teams prioritize issues with hotel listings and trained them to use it, increasing the number of critical tasks resolved by 25%.
  • Used SQL and language processing to find inconsistent room types in listing descriptions, improving detection by 120% compared with the existing rules.

Selected work.

Academic and personal projects in AI and machine learning. Explore the code where it’s available.

  1. Personal project · MTA Open Data · 2026

    Where Does New York Go?

    Built an interactive origin-destination flowprint that reveals how arrivals, departures, and station connections change across the day in New York City.

    MTA Open Datadeck.glReact
    Explore project
  2. Kalamata Capital Group × Columbia · 2025

    Credit Risk Modeling

    Developed models to assess credit risk using 100,000 transactions, tested them on later time periods to check how well they generalized, and used explanations to show which factors influenced their predictions.

    CatBoost / XGBoostTimeSeriesSplitSHAP
    Code not public

My education.

Sep 2025–Dec 2026 (expected)

Columbia University

M.S. in Operations Research

Applied Machine Learning, Optimization Models & Methods, Agentic AI, and Analytics on the Cloud.

Aug 2019–Oct 2023

Chulalongkorn University

B.Eng. in Robotics & AI Engineering

First Class Honors · Academic Excellence Scholarship (2022) · RoboCup, 2nd place.

Technical toolkit.

Data & pipelines
Python, SQL, Pandas, NumPy, Spark SQL, PySpark, Trino, Airflow, DuckDB
Machine learning & optimization
Scikit-learn, XGBoost, CatBoost, BERT / RoBERTa, SHAP, vehicle routing, mixed-integer programming
AI applications
OpenAI Agents SDK, AWS Bedrock, LangChain, LiteLLM, RAG, FAISS, output validation
Shipping & communicating
Docker, Google Cloud Run, CI/CD, FastAPI, React, Tableau, Power BI, Kepler.gl