When we first invested in Spangle AI, we knew Maju and the team had a clear vision for agentic systems in commerce, and we saw strong early value in the product itself. What we did not fully anticipate was how quickly and deeply customers would resonate with Spangle’s technology and pull the product into broader use.
This is an indicator of growing interest in agentic systems across consumer brands. The premise is straightforward: software that can respond to customer intent, take action in real time, and improve with use should materially improve conversion and efficiency. Many leading brands see this future coming, but delivering real solutions and sustained impact has been rare.
Spangle has been a clear exception. Today, we’re pleased to double down and participate in their Series A.
Since the seed round, the most meaningful change has been customer excitement. Brands are asking Spangle to show up in more places and support an even wider range of use cases. What began with ad-driven traffic has expanded into email, onsite experiences, and increasingly into answer engines and AI-led discovery surfaces like ChatGPT. The surface area Spangle is being asked to cover has grown significantly, driven by customer demand even more than roadmap ambition.
That level of pull is difficult to manufacture and hard to ignore. It is a strong signal that the underlying system is solving a real problem and earning trust quickly.
Where Commerce Breaks Down
The intensity of that customer pull reflects a real shift in how commerce is operating.
Customer intent increasingly forms upstream of a brand’s website. Search, social platforms, recommendation engines, email, and AI‑driven interfaces shape decisions before a shopper ever lands on a product page. Brands are feeling this change directly, as performance depends on how well intent carries across surfaces.
The systems responsible for conversion were not built for this environment. Data is fragmented across channels. Feedback cycles are slow. Insights arrive after the opportunity has passed. Teams spend more time optimizing reports than improving outcomes.
As brands push Spangle into new surfaces, this limitation becomes more visible. Any system that claims to be agentic in commerce has to operate across these environments and still deliver measurable results.
How Spangle Approached the Problem
Spangle did not start with features. They started with the future. We have known Spangle’s founder, Maju Kuruvilla, for years before he started Spangle, and have always been impressed by his vision for the future of customer-centric brand relationships.
From the beginning, the Spangle platform was designed to connect intent to execution regardless of where that intent originates. That architectural choice is why customers have been able to extend Spangle from paid traffic into email, onsite experiences, and now AI‑led discovery surfaces.
At the center of the platform is ProductGPT, a commerce‑specific reasoning engine that models a brand’s products, context, and conversion behavior. Seller Agents operate in real time, adjusting experiences and executing decisions as shopper intent evolves across channels. Results flow back into the system and inform future interactions.
The important point is not the individual components, but the loop they create. Action produces signal. Signal improves the model. The system gets better with use.
Learning systems improving with scale is not new in commerce. Traditional machine‑learning‑driven optimization has benefited from more data and more traffic for years. What is different here is the nature of the learning itself.
Spangle’s system reasons in real time, acts directly in the customer experience, and incorporates outcomes immediately. That tighter loop enables faster learning and more effective optimization as the surface area expands.
This is what makes Spangle core infrastructure, not mere tooling.
Why This Team Has Earned Conviction
We partnered with Spangle early because they understand both sides of the problem. Maju and his co-founder, Fei Wang, have built AI systems in environments where scale, reliability, and economics matter. They understand how learning systems behave under real traffic, and what it takes to make them perform consistently as conditions change.
That experience is particularly important as customers pull the platform into broader use. Supporting more surfaces without fragmenting intelligence requires discipline at the system level.
That discipline shows up in execution. In under a year, Spangle has driven meaningful improvements in conversion and return on ad spend for a fast-growing roster of world class brands. Just as importantly, those gains have continued as traffic volumes and use cases expand.
Why We Doubled Down
We believe agentic systems will become part of the core infrastructure for consumer brands. The systems that matter will be those that can operate across discovery, engagement, and conversion, and learn in real time as conditions change.
Spangle is focused on the hardest part of the problem: building a learning system that connects intent to execution and scales with customer demand.
The work ahead is substantial, but the foundation, vision, and momentum are strong. We look forward to continuing the partnership and seeing more brands benefit from agentic AI.
Related Insights
