Profile
Soichiro works across application development, visual impression AI model development, design, and content production. He is currently planning, designing, and developing CheckIn, a safety confirmation app; Spotgram, an anime location pilgrimage and local discovery app; and Mehrabian, an AI research project for evaluating visual impressions of human images.
His app development covers more than UI implementation: algorithms based on user behavior and daily rhythms, privacy-conscious data design, subscription models, onboarding design, and App Store visual production.
He also produces video and narration and creates educational podcasts, bringing technology, design, and language together in product development.
Projects
Spotgram
Spotgram is an anime location pilgrimage app that allows Users to post and share locations featured in anime and other works together with music. Users can record the locations they have visited and share music and memories associated with those places.
Going beyond a map app, Spotgram aims to be a local discovery platform that combines anime, music, tourism, local culture, and user posts to help people encounter new places through the works they love.
CheckIn
CheckIn is a safety confirmation app that quietly helps confirm the wellbeing of important people based on everyday smartphone use. Rather than requiring a manual report each time, it is designed to record check-in information automatically using iPhone Shortcuts and routine interactions.
The initial specification provides a simple monitoring feature that sends a notification when no check-in has been recorded for a set period. Development is progressing on an algorithm that detects unusual situations more naturally based on each User's rhythm and usage patterns.
It is designed not merely as a notification app, but as a product that uses everyday behavioral patterns as signals of safety while respecting privacy.
MehrabianAI
Mehrabian is an independently initiated large-scale evaluation model research and development project. It aims to estimate human visual impressions and preferences for images of people based on many human evaluations, in a manner closer to real-world evaluation tendencies. Rather than simply scoring images, it works toward a stable model that accounts for individual differences, evaluator bias, comparison conditions, and skewed data distributions.
Its research includes pairwise comparisons, rating models, Bradley-Terry models, Thurstone-Mosteller models, hierarchical Bayesian models, normalization and standardized scoring, and labeling tool design. Building on these methods, the project develops an evaluation framework suited to subjective judgments of human images.
Future work aims to integrate local labeling tools, multi-person evaluation data collection, server-based data management, AI inference models, and score calibration into an experimental platform for visual impression assessment and practical deployment.
Podcast: Chie no Mori
Chie no Mori is an audio program covering science, technology, history, companies, and social structures. More than 30 episodes intended to stimulate intellectual curiosity have been released, with planning, structure, scripts, narration, and distribution all produced by Soichiro.
It aims to translate specialized subjects into a form that general listeners can understand and deliver them as compelling narratives.
For Investors and Business Partners
Development is underway on multiple products centered on CheckIn, Spotgram, and MehrabianAI, focusing on daily behavioral data, local experiences, and visual impression AI.
Each project addresses areas closely tied to everyday human life, such as habits, movement, memories, emotions, and evaluative behavior. The goal is to develop them beyond individual apps or tools into a continuously growing group of products combining data infrastructure, subscription models, regional collaboration, and AI model development.
While currently based on independent development, the work spans planning, design, development, content production, and product strategy. Dialogues are welcome with investors and business partners regarding business development, funding, joint research, partnerships with local communities and companies, and cooperation in AI and data utilization.
For inquiries, please use the contact button below.