Behavior intelligence startup Malachyte raised $10 million in a seed funding round co-led by Bessemer Venture Partners and Gradient Ventures to expand its real time, AI driven e-commerce personalization platform.
Malachyte, a New York-based behavior intelligence company, closed a $10 million seed funding round. The round was co-led by Bessemer Venture Partners and Gradient Ventures, with participation from Harpoon Ventures. The capital will primarily support scaling distribution of the existing platform and hiring senior commercial and product leadership.
What is Malachyte?
Malachyte was founded in 2024 by Sidd Motwani (CEO), Ian Anderson (CTO), and Shivaditya Sinha (COO). The co-founders previously built the behavioral intelligence infrastructure (known as Vector AI) that powers more than 90% of Spotify’s recommendations across 800 million+ users and 1 billion+ items. Motwani has over 12 years of experience building data driven products at Priceline and Spotify, Anderson served as a Staff ML Engineer at Spotify (leading teams of 100+ and helping redesign core recommendation systems while serving as Founding Chair of Spotify’s AI Advisory Board), and Sinha brings venture capital, consulting (BCG), and operating experience, along with an MBA from Columbia Business School and London Business School.
The company focuses on real time personalization for consumer e-commerce brands. It aims to increase revenue per visitor for every shopper across every channel, starting from the first session, without relying on cookies, logins, or pre existing customer data.
E-commerce faces a conversion challenge. Advances in machine learning driven ad targeting on platforms like Google, Meta, and TikTok have improved audience acquisition precision, but most website experiences remain static. Traditional personalization stacks struggle to interpret shopper context, adapt in real time, or effectively serve first time (cold start) visitors. Customer acquisition costs rose roughly 40% between 2023 and 2025, and the average e-commerce brand loses about $29 per new customer after marketing and returns. Unconverted visits compound these losses.
Most systems rely on historical purchases, demographic segments, or logged-in profiles, leaving anonymous or first time visitors with generic experiences and failing to act on live session signals.
What is Malachyte’s technology?
Malachyte’s core technology is a two headed vector AI architecture that simultaneously infers shopper preference (general taste) and intent (what the shopper is trying to accomplish right now). It builds adaptive user profiles from behavioral signals (clicks, searches, hovers, scrolls, browsing context, device, time of day, referral source), visuals, and context, starting before the first click and continuously refining them.
Key capabilities include:
- Continuous, real time learning rather than batch (nightly/weekly) retraining.
- Cold start handling via representation fusion in seconds, without needing purchase history or logins.
- Cross surface persistence (signals from search influence product pages, category pages, email, or SMS later).
- Production performance: sub 200 millisecond latency under peak Cyber Monday level loads, auto scaling, and no reported outages.
- Merchandiser control layer allowing operators to set rules and overrides.
- Fast onboarding (as little as seven days).
- Native Shopify integration (available since June 2026) and API access for larger retailers.
The platform first went live in fall 2025 with Fun.com (HalloweenCostumes.com). Prior to focusing on e-commerce, the team tested the technology with more than 20 enterprise customers across travel, grocery, and retail since 2024.
Recommended: Lydian Raises $43 Million In Series A Funding
Early Customer Results
Reported performance metrics from live deployments include:
- HalloweenCostumes.com (Fun.com): +31% revenue per visitor.
- Brunt Workwear: 80% lift in add to cart click through rate.
- Jordan Craig: 17% lift in new visitor revenue per visit (notable because ~80% of traffic consists of first time visitors with no purchase history).
Customer feedback highlights the ability to move from guessing to showing relevant products in the moment and improving session value on product detail pages.
The funding arrives amid rising customer acquisition costs and the emergence of agentic commerce. McKinsey has projected that agents could orchestrate up to $1 trillion of U.S. retail by 2030. Bessemer’s Maha Malik noted that consumers show less patience for non tailored experiences while agents increasingly shape buyer journeys, creating demand for real time intent understanding that most existing commerce infrastructure lacks.
Malachyte positions its behavior intelligence as infrastructure for both human shoppers and future AI agents, aiming to unify merchandising and marketing around a shared, real time understanding of behavior. The approach emphasizes privacy by design (no reliance on identity data or third party cookies) and transparency for brands.
The $10 million will fund expanded go to market efforts and recruitment of senior commercial and product talent. The company is already live with a growing set of brands and continues to sign new retailers. Dealroom data places estimated enterprise value in the $40–60 million range following the round.
The round validates the transfer of proven large scale recommendation infrastructure from Spotify to e-commerce personalization, targeting a clear gap in real time, cold start capable systems as online retail faces higher acquisition costs and evolving shopper (and agent) expectations.
Please email us your feedback and news tips at hello(at)techcompanynews.com
