Building quiet systems with visible craft.

Developed consumer-facing product features across discovery, checkout, and post-order experiences for a production commerce platform.

• Unified fragmented consumer app codebases into a single cross-platform foundation.
• Improved development velocity by introducing clearer architecture boundaries and reusable abstractions.
• Tuned critical consumer flows with performance measurement and iteration.

Built a social media intelligence pipeline that collected, modeled, and served large-scale landscape and trend analysis across X, Instagram, LinkedIn, Facebook, and TikTok.

• Integrated direct platform collection and Apify-based workflows into dashboard-ready data products.
• Transformed social media streams into episodic topic modeling so core narratives could be tracked historically across initiators, timelines, and subtopics.
• Built continuous topic modeling workflows with BigQuery, Transformers, Sentence Transformers, BERTopic, and Python scraping tools.
• Optimized analysis cost by partitioning BigQuery tables around social media timelines and query patterns.
• Built alerting to detect emerging narratives before they became trends, supporting faster and more useful business decisions.
• Mapped sentiment, bots, buzzers, and social signals into action items so internal teams could respond with cleaner data.

Mentored Digital Talent Scholarship participants through a focused frontend development workshop.

• Designed practical frontend learning modules for Digital Talent Scholarship participants.
• Led 10+ live mentoring sessions focused on React and project execution.
• Coached participants from implementation through final project assessment.

Built frontend user and admin applications for a supply-chain MOOC product.

• Built learner and admin product surfaces for a supply-chain education platform.
• Connected frontend workflows to REST and GraphQL-backed data models.
• Integrated Midtrans payments into the enrollment and transaction flow.

Worked on sentiment analysis and recommendation system experiments for Indonesian text data.

• Prototyped Indonesian-language sentiment analysis and recommendation workflows.
• Built text preprocessing pipelines for normalization, tagging, embedding, and weighting.
• Implemented model experiments using Keras-based recurrent neural networks.

Built mobile frontend applications for a workflow management system across Android and iOS.

• Built Android and iOS workflow-management surfaces with React Native.
• Connected mobile screens to REST and GraphQL-backed data flows.
• Helped establish repeatable mobile build workflows.