From Digital Libraries to Decentralized Insurance
Bill Liu graduated from the University of North Carolina at Chapel Hill with a degree in Library and Information Science and Philosophy - a pairing that sounds like a punchline until you realize he spent his twenties thinking about how information flows, who controls it, and what happens when systems fail the people they're supposed to serve.
Before Seel, before pitch decks and term sheets, Liu was writing code for ibiblio.org - one of the internet's oldest public digital libraries - and running IT operations for UNC's research computing center. Both jobs have a common thread: infrastructure work. The unglamorous, foundational kind where you keep systems running so that other people can do their thing without thinking about you at all.
That sensibility - build the layer beneath the layer - would become Seel's entire business model.
Insurance is an important societal infrastructure. We have a unique opportunity to fix this fundamental yet obsolete social infrastructure.
- Zack Peng, Co-Founder & CEO, Seel (on the founding thesis)In 2019, Liu and co-founder Zack Peng - a former underwriting engineer who had spent years inside the insurance machine - applied to the MetLife Digital Accelerator powered by Techstars. The program was looking for startups that could disrupt insurance from the outside. Liu and Peng were building something called Kover.ai, a pitch to reinvent insurance infrastructure using smart contracts and decentralized pools of capital. In other words: replace the middlemen with math.
They made it into the cohort. Nine companies got in that year. Very few pivoted as dramatically, or as successfully, as Kover.ai would.
When "Decentralized Insurance" Meets a Shopping Cart
The Problem Hiding in Plain Sight
Ecommerce returns cost retailers 8% of store margins on average. Merchants don't know their true revenue until the return window expires six to eight weeks after an order ships. Shoppers abandon carts because they can't return final-sale items. Everybody knows it's broken. Nobody had figured out how to sell the fix at checkout.
The original Kover.ai proposition - decentralized insurance rails for any business - was bold and technically sound. But the market kept asking a narrower question: can you solve the return problem, specifically? Returns were keeping ecommerce merchants up at night in a way that nothing else quite was.
Peng and Liu listened. They sharpened the product focus and rebranded as Seel. The pivot wasn't a retreat. It was a better aim at a bigger, more immediate problem: the $816 billion in merchandise returned globally every year, much of it handled by manual, margin-destroying processes that no one had thought to automate with AI.
The company launched its private beta in mid-2021. Within five months, roughly 200 merchants had signed up. One in four shoppers shown the offer bought Seel's Return Assurance at checkout. Conversion rates lifted 5%. Merchants liked it so much that Seel's word-of-mouth was doing significant GTM work before the company had formally launched.
An AI That Prices Risk Before the Package Ships
Seel's core product, Return Assurance, is elegant in a way that few insurtech ideas manage to be. A shopper adds a pair of shoes to their cart from a final-sale brand. At checkout, Seel appears as an option - add return coverage for a small fee. If the shoes don't fit, Seel refunds the purchase directly. The merchant never sees the return chargeback. The shopper gets peace of mind. The risk is absorbed by a system that knows, before the box is sealed, how likely that return is.
The pricing engine is a neural network that runs on hundreds of signals simultaneously - product category, purchase history, shopper behavior patterns, order characteristics - to calculate return probability at the exact moment an order is placed. The algorithm doesn't just predict. It prices. Every risk assessment is also a pricing decision.
This is the kind of infrastructure that sounds boring from the outside and represents an enormous technical moat from the inside. Seel's database of aggregated ecommerce return data, trained across thousands of merchants and tens of millions of orders, is not something a competitor can spin up in a quarter.
Bill Liu's Job: Build the Machine That Runs the Machine
While Zack Peng (CEO) tends to the product vision and Mo Chen (CTO, former Google engineer) runs the technical architecture, Liu's charter has been the harder-to-publicize work: building teams, driving go-to-market, and scaling operations in a company that grew from five employees to 150+ in a few years while processing millions of real insurance transactions.
Liu's LinkedIn URL still carries his Chinese name - zhangsinong-liu - a small tell that he operates without a lot of performative self-branding. He shares Seel's CB Insights coverage. He posts about the Seel playbook. He does not post about himself.
That tracks with what COOs actually do: they build the organization that makes the vision real. At Seel, that meant hiring across merchant success, risk operations, partnerships, and data science - the functional breadth that a company needs when it's simultaneously an insurance company, a SaaS platform, and a merchant-facing ecommerce tool.
Seel's Return Assurance lets shoppers add a 7-day refund window when buying final sale items - helping 500+ leading brands deliver risk-free shopping while adding 30% net new revenue.
- Seel company overviewThe go-to-market playbook Liu helped design is worth studying. Seel integrates directly into Shopify, Magento, and WooCommerce checkout flows. Merchants don't rebuild anything. The shopper experience stays native. The data flows back to Seel's risk engine. It's a wedge strategy - low friction entry, high switching cost over time as the merchant's return data becomes part of Seel's model.
By 2024, the platform had grown to over 5,000 merchants and had handled more than 24 million protected orders. Authentic Brands Group, Debenhams, and Shoplazza had all signed on as partners. Seel won Shoplazza's 2024 Partner of the Year award.
From $1.5M Seed to a Series B
The money trail is a useful map of Seel's credibility arc. The seed round in November 2019 - $1.5 million, led by West Loop Ventures, with Afore Capital and Techstars Venture - was a bet on two founders and a thesis about decentralized insurance infrastructure. At that point Kover.ai was still figuring out what to be.
The Series A in January 2022 - $17 million, led by Lightspeed Venture Partners, with Foundation Capital, Afore Capital, and West Loop Ventures returning - was a bet on proof. Seel had merchants. It had data. It had a 1-in-4 opt-in rate that most consumer fintech products never achieve.
The Series B, which closed in May 2025, brought total funding to $29 million-plus. West Loop Ventures, a fund that has been in since the very first check, remains a backer. That kind of investor continuity across three rounds is its own signal.
The Specifics
Seel started as Kover.ai - a decentralized insurance play using smart contracts. The pivot to ecommerce return assurance was not an obvious move, and it worked.
Bill Liu's LinkedIn handle is "zhangsinong-liu" - a nod to his full Chinese name, quietly present in a profile that otherwise runs lean on personal detail.
Liu's academic background in Library and Information Science is not the typical founder resume - it's a systems-thinker's education, and it shows in how Seel was built.
The Seel neural network prices return risk using hundreds of signals per order - in the time it takes a shopper to click "add to cart," the model has already run its assessment.
Seel's repeat shopper opt-in rate is even higher than the first-time rate - meaning the product builds its own loyalty loop among consumers who've used it.
West Loop Ventures backed Seel at seed, Series A, and Series B - three straight rounds from the same lead, which is unusual and tells you something about what they see in the founders.