Price: 0
Number of applications: 8
27.08.26 (inclusive)
Contractual, based on the results of consideration of the proposed solution
MVP
ICT tasks
Media sphere
Information processing and transformation
Software/ IS,
Mobile app
DEAWA is developing an AI personal recommendation service for free time. The system can offer the user places, activities, home activities, content, and other leisure options. For a single user request, the system can receive dozens of potentially suitable options from different sources. Simple sorting by rating or distance does not allow you to take into account the individual context of the user. It is necessary to develop a technology module that will automatically evaluate and rank the available options, taking into account several factors at the same time.: current location; time and available duration; budget; user interests; previous likes and dislikes; recommendations already shown; preferences for active or relaxing holidays; Vacation formats: single, couple, friends, family; distance and travel time; the opening hours of the facility or the start of the event; weather conditions; matching the user's current request; preferences for familiar or new types of activity. The problem is the need to combine these parameters into a single ranking mechanism that will work quickly, predictably, and can gradually improve based on user feedback.
The development of the personal ranking module will increase the relevance of IDEAWA recommendations by simultaneously taking into account the user's interests, current context, budget, distance, time, interaction history and feedback. It is expected to reduce the number of irrelevant and repetitive recommendations, increase the variety of suggested options, and increase the proportion of recommendations that the user interacts with. The solution should also create a technological basis for the subsequent transition from rule-based ranking to an ML/AI model trained on accumulated user feedback.
Boldyshevsky Dmitry Ilyich
Purpose and description of task (project)
To develop a prototype of a personal recommendation ranking module that receives a set of prepared objects and a user context, calculates the relevance of each option, and returns a sorted list of the most appropriate recommendations.
Note
IDEAWA is a mobile application with AI personalization of leisure time. The project is at the MVP stage. As part of this task, proposals for the architecture and implementation of the personal ranking module are being considered. The preferred approach is to start with a deterministic scoring model and then move on to ML/Learning-to-Rank based on accumulated user feedback. Anonymized test data, the structure of the input JSON, examples of recommendation objects, and user test scenarios can be provided to evaluate the solution. API keys, user personal data, and production access are not provided at the initial review stage.
Whatsapp
