Malachyte, a startup founded by three former Spotify engineers who built the recommendation system that drives roughly 90% of Spotify's suggestions to its 800 million users, has raised $10 million in seed funding to port that approach to online retail. Bessemer Venture Partners and Gradient co-led the round, with Harpoon Ventures participating. The company will use the capital to scale distribution and hire product and commercial leaders.
Co-founders Sidd Motwani, Ian Anderson, and Shivaditya Sinha spent years at Spotify developing Vector AI, a system that predicts a listener's next action rather than merely recommending against past behavior. Malachyte is now selling a version of the same idea to e-commerce merchants, wrapped as what it calls a "two-headed Vector AI" that predicts the next product a shopper wants while continuously refining a taste profile in-session.
The pitch is that most online stores personalize badly. First-time visitors see a generic storefront, and returning customers get recommendations built on what they bought last year rather than what they came for today. Malachyte's system starts inferring intent the moment a page loads, using contextual signals — device, referral source, time of day — before a shopper has clicked anything.
“[Our] system starts forming before the first click, using the context available the moment the page loads. Within a single session, we build a real read on both preferences and what someone is trying to accomplish right now.”— Sidd Motwani, Malachyte CEO
Key facts
- 01Malachyte raised $10M in seed funding co-led by Bessemer Venture Partners and Gradient, with Harpoon Ventures participating.
- 02The founding team previously built Spotify's Vector AI, which drives roughly 90% of recommendations for its 800M users.
- 03The platform launched with Fun.com in fall 2025 and reached general availability on Shopify in June 2026.
- 04Malachyte tested with more than 20 enterprise customers across travel, grocery, and retail before narrowing focus to e-commerce.
Motwani gave the example of a search for "heavy-duty boot" followed by two clicks on steel-toed boots. Within seconds, the storefront moves work pants and gloves up the page and pushes dress shoes down, with no account or purchase history required. Every subsequent hover, scroll, or add-to-cart sharpens the model further, and Motwani says the effect compounds across visits.
That real-time signal loop is the part Motwani argues most retailers get wrong. Existing personalization engines tend to batch behavioral data into overnight segments, meaning the storefront a shopper sees in one session was tuned to yesterday's aggregate cohort. Malachyte reads each action as it happens and updates the user's vector on the fly.
Contextual state matters too. An email-driven mobile visit at 11 p.m. represents a very different shopping mindset than a laptop session mid-morning, and Motwani says most e-commerce stacks fail to distinguish between the two even when they have the raw data.
Malachyte began developing and testing its technology in 2024. It worked with more than 20 enterprise customers spanning travel, grocery, and retail before narrowing its focus to e-commerce specifically. The platform first went live with Fun.com in fall 2025 and reached general availability on Shopify in June 2026 through a native integration. Larger retailers can plug in via API.
Motwani says the next expansion is joining merchandising and marketing around a shared behavioral model — meaning the same real-time intent signal that reranks a product grid also drives the email, the ad retarget, and the on-site promotion. In practice, most retailers today run those systems on separate data pipelines and separate cadences.
The pitch faces real competitive pressure. Shopify itself has been investing in native AI-driven merchandising, and every major e-commerce platform vendor has shipped some flavor of personalization engine in the past two years. Malachyte's differentiation rests on whether the Spotify-trained team can genuinely deliver a step-change in session-level relevance versus incumbent tools that merchants already pay for. The 20-customer pilot base is a start, but the Shopify app store is a crowded room.
The interesting bet inside this round is that recommendation infrastructure is becoming a horizontal category. The techniques that made Spotify's home feed work — vector embeddings of user state, continuous online learning, contextual bandits — are increasingly viable for any business with a product catalog and a session log. If Malachyte can convert its Shopify beachhead into meaningful conversion lift, expect a wave of consumer-AI alumni pitching similar spin-outs into every vertical with a checkout button.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
