ShopMelody
Online shoppers often browse e-commerce websites without the immersive atmosphere of physical stores, where background music helps create a mood and make shopping more enjoyable. This is especially relevant for fashion, beauty, and lifestyle shoppers who rely on visuals and product descriptions but lack an engaging sensory experience online. At the same time, e-commerce businesses struggle to create emotional connections with customers and stand out in a competitive market. ShopMelody addresses this gap by using AI to automatically recommend and adapt background music based on the products customers browse, making online shopping more personalized, engaging, and memorable.
Preliminary score
- Value for the user8.5 / 25
A specific problem and a clear outcome from solving it.
- Prototype and use of AI2.0 / 30
A working core scenario, and what AI actually contributes to the result.
- Quality testing2.5 / 20
Tests, examples of failures, and comparison with the current approach.
- Feasibility3.5 / 15
Data requirements, running costs and a clear next step.
- Originality4.5 / 10
A meaningful difference from common solutions.
What the judges wrote
- GPT-6 AstraPreliminary17/100
- Strengths
- Your proposal offers a sensible experiential twist: adapting music to products browsed in fashion, beauty, and lifestyle stores. Your submission clearly distinguishes the engagement hypothesis from proven results and candidly acknowledges the absence of a functional prototype and user tests.
- To improve
- Your next build should demonstrate one complete flow from MyBrands product metadata through AI-based music selection to opt-in playback; the supplied materials do not establish a working AI scenario. Your testing should compare adaptive music against silent browsing using predefined measures such as shopping satisfaction, purchase completion, and mute rates, while documenting distracting or irrelevant selections. Your feasibility plan should cover the music source and licensing, required product data, playback restrictions, and estimated running costs.
- Claude Opus 5.5Preliminary25/100
- Strengths
- Your write-up is refreshingly honest: you clearly state that no models are used yet and that no user tests have been run, rather than overclaiming. You also name the right risks to test (music relevance, seamless playback, user preferences, avoiding distraction) and propose a sensible A/B comparison of engagement with and without adaptive music. Tying music to the product being browsed via the MyBrands API is a recognisable twist on site personalisation.
- To improve
- Your materials show no working AI: the README is the default React + Vite template and no model is involved, so the core scenario (product → music selection) needs to actually run, even as a simple LLM or embedding mapping from product attributes to mood or genre playlists. For the Enterprise track, pick one metric a retailer would care about (time on page, add-to-cart rate, bounce rate) and run even a small within-team or hallway test with a few concrete examples, including failures. Address music licensing costs, autoplay restrictions in browsers, and API or running costs, because these decide whether a shop could adopt this.
The team
Leyla AhmadovaLead
developer