Kensuke
Umakoshi
Data & AI Professional. Building intelligent systems from Cloud to Code. Passionate about solving real-world problems using data, from cloud-native ETL pipelines to interactive dashboards.
End-to-end data engineering platform with AI SQL generation and analytics.
AI-powered invoice processing pipeline with duplicate detection and automated summaries.
AI-powered agent handling inquiries with policy-based responses and intelligent escalation.
Multi-level ticket classification system using BERT-based models to categorize service tickets with 97% accuracy.
Interactive dashboard visualizing H-1B visa data to help international students find sponsorship opportunities.
Automated CI/CD pipeline using AWS S3, Terraform, and GitHub Actions for seamless website deployment.
Real-time fraud detection system built with a scalable ML pipeline, Docker, and API integration.
Predictive model analyzing healthcare data to identify trends and optimize Medicare resource allocation.
Personalized movie recommendation engine combining collaborative filtering and content-based NLP techniques.
Automated news aggregation and analysis tool using LLMs to extract key insights from global headlines.
Intelligent chatbot designed to assist students with study schedules, resource finding, and Q&A.
Comprehensive SQL and Python analysis of sales data to drive business insights for a food delivery platform.
Data-driven analysis identifying key factors contributing to customer churn in the banking sector.
Teaching Assistant, Intro to Business Analytics
University of Utah | Salt Lake City, UT- Mentor students in Python programming and machine learning concepts.
- Facilitate clear communication of technical results to help students develop strong analytical and data storytelling skills.
Data Engineering & Analytics Project
U Career Success, University of Utah | Salt Lake City, UT- Drove end-to-end development of university-wide visa sponsor job dashboard serving 1,000+ international students: gathered requirements from career coaches, built automated quarterly ETL pipeline processing 240,000+ H-1B records (Python), and designed interactive Tableau visualizations with company, location, job type, and visa filters.
- Collaborated cross-functionally with career advisors and IT team to deploy and integrate dashboard into university career services website, replacing decentralized manual research with centralized H-1B Employer Data Hub access and streamlining sponsor search process.
Customer Insights Analyst
Cortex Evolution | Salt Lake City, UT- Analyzed 200+ customer behavior records using Google Analytics and Hotjar to elevate user engagement.
- Developed interactive Tableau dashboards to support campaign decisions.
- Automated Reddit thread scraping via API to identify potential market leads.