Technology news and discussion on TalkFuse.
Ex-Spotify employees raise $10M to bring the AI behind its recommendations to e-commerce
Three former Spotify engineers who built the streaming service's Vector AI recommendation system have raised $10 million in seed funding for Malachyte, a startup applying the same technology to online shopping. The company argues that most e-commerce sites still personalise based on stale purchase history or demographic data, leaving first-time visitors with generic storefronts and returning customers with recommendations that ignore what they actually want right now.
Malachyte's "two-headed Vector AI" builds a real-time profile of a shopper's intent from the moment a page loads, using signals such as clicks, hovers, searches and device context to reorder products within a single session, without requiring an account or purchase history. Founded by Sidd Motwani, Ian Anderson and Shivaditya Sinha, the company has tested its technology since 2024 with over 20 enterprise clients, launched with Fun.com in autumn 2025, and has been generally available to Shopify merchants since June 2026, with larger retailers integrating via API. The seed round was co-led by Bessemer Venture Partners and Gradient, with participation from Harpoon Ventures, and will fund distribution and hiring.
In short
- Ex-Spotify engineers raise $10M for Malachyte, an e-commerce AI startup
- Uses "Vector AI" to predict shopper intent in real time, no login needed
- Live with Fun.com and Shopify merchants; funding led by Bessemer and Gradient
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Malachyte's approach represents a genuine improvement over existing e-commerce personalisation, which relies on outdated purchase history or crude demographic data that provides poor results for first-time visitors and returning customers alike. By analysing real-time behavioural signals—clicks, hovers, searches—during a single session, the system tailors recommendations to what shoppers actually want in the moment without requiring account creation or historical data. This respects user autonomy, improves discovery for legitimate retailers, and levels the playing field so smaller merchants can compete with giants through better product matching rather than advertising spend.
The case against
The collection and processing of granular behavioural data—every click, hover, and search—raises serious privacy concerns, particularly since users may not explicitly consent or understand they're being profiled in real time. Such detailed tracking enables algorithmic manipulation through dark patterns and personalised pricing strategies, whilst creating filter bubbles that steer consumers toward high-margin products rather than what they genuinely need. The concentration of recommendation power in AI systems run by a small number of companies reduces transparency and consumer agency, particularly when many users cannot easily opt out and alternatives remain limited.
Explained simply
Online shops often personalise what shoppers see using past purchases or demographic information. That can make recommendations less relevant for new visitors, or when someone’s interests change during a visit.
Malachyte is a startup founded by three former Spotify engineers: Sidd Motwani, Ian Anderson and Shivaditya Sinha. They are adapting recommendation technology they worked on at Spotify, where algorithms suggest music, to help online retailers choose which products to show each shopper.
The technology aims to infer a shopper’s current interests from actions such as searches and clicks, without needing an account or purchase history. It matters to retailers because more relevant product displays could help shoppers find what they want, while raising questions about how browsing signals are used.
AI Art Business Celebrity Companies Culture Entertainment Music Technology