Rayhan Titho

Led MAP Group’s Central Data & AI efforts to build a centralized Databricks Lakehouse platform on AWS under a hub-and-spoke operating model across multiple brands and business units. Architected causal-inference-based discount recommendations and optimized Databricks/Unity Catalog governance plus Tableau reporting performance and monitoring.

Developed machine learning models for moisture detection, maintenance fraud anomaly detection, and predictive maintenance to improve operational reliability and planning. Built safety and operational analytics dashboards, including production loss/capacity metrics and truck operator performance tracking.

Built an AI chat-based flowchart builder and a voice agent to automate logistics workflows and supported product demo engineering for client-facing demos.

Mentored university students in Python, Pandas, SQL, and end-to-end ML project design, supervising final capstone projects.

Built a music similarity search model using audio feature embeddings and triplet-loss training, evaluating performance with ranking metrics. Deployed an inference pipeline using pgvector and improved matching with Whisper STT, and automated data workflows including BigQuery migration and multimodal LLM-based ingestion.

Built Kredivo’s first audio ML model to classify voicemails, preventing significant auditor costs. Applied cohort and survival analysis on a delinquent portfolio to determine write-off timing thresholds.

Built and deployed the company’s first recommendation system using user similarity and association rules, improving add-to-cart and checkout conversion. Transitioned an auto driver routing deployment from manual Jenkins triggering to a production API approach.

Developed semantic role labeling and sentiment classification models with improved word-level explainability for media clients. Built an automated labeling system to increase throughput, upgraded the sentiment model from GloVe to BERT, and used Weights & Biases for experiment tracking.

Grade: GPA 3.52/4.0

Earned an M.Eng in Information Technology with a thesis on music similarity ranking, benchmarking audio similarity algorithms using Mean Rank, MRR, and MAP.

Grade: GPA 3.25/4.0

Earned a B.Eng in Electrical Engineering with a thesis on robotic localization, implementing and benchmarking AMCL vs. MCL on the TurtleBot3 platform.