| Title | TASTECH ENGINE |
| Brand | NONGRID |
| Product / Service | TASTECH ENGINE |
| Category | D02. Data Storytelling |
| Entrant | HAKUHODO INC. Tokyo, JAPAN |
| Idea Creation | HAKUHODO INC. Tokyo, JAPAN |
| Production | NON-GRID Tokyo, JAPAN |
| Name | Company | Position |
|---|---|---|
| Hiroyuki Shinoda | HAKUHODO | Data Scientist |
| Yusei Kakizaki | HAKUHODO | Art Director |
| Shohei Kurita | HAKUHODO | R&D Planner |
| Daisuke Niino | HAKUHODO | Interactive Planner |
| Aya Suda | HAKUHODO DY Digital | Data Analyst |
| Koki James Sase | FREELANCE | Planner |
| Chako Kibe | NON−GRID | Why Juice? Director |
| Aiko Yaguchi | NON−GRID | Producer |
| Natsuko Miyazaki | IMG SRC | Designer |
| Masanori Yoshii | IMG SRC | Technical Director |
Your preference for food changes with timing and your emotion. And it is difficult to search what you really want to eat from vast amounts of information of recipe or restaurant site because many of them you have not eaten yet. So, we create "TASTECH ENGINE" which estimates your real-time taste preferences using big data.
We analyze the 100 million records of browser data focused on recipes and other food sites, and create "TASTECH ENGINE" which estimates your real-time taste preferences “right now” using big data. We introduced TASTECH ENGINE on a juice bar's website.When a user access the site, their most recent recipe site browsing trends are automatically used to predict their preferences, and a recommended juice recipe is displayed.If they show that recipe screen at the juice bar, they can get their own original juice mixed for them. So far this service has continuously estimated real-time eating preferences and dietary habits of more than 2 million consumers online.33% of site visitors learned what their own tastes were.As a result, juice bar sales went up 2.5%.