Source: CYZONE (Chuangyebang)

On January 13, CYZONE exclusively learned that Seel (formerly Kover), an AI-powered insurance company focused on the North American market, has recently completed a $17 million Series A funding round. This round was led by Lightspeed Venture Partners and marks the largest Series A funding in the U.S. insurance actuarial space.

The funding will primarily be used for team expansion. Seel is headquartered in San Francisco, with an additional office in Salt Lake City, Utah. This year, Seel plans to establish a research and development center in Shanghai, aiming to recruit back-end engineers capable of helping build its actuarial engine.

Seel develops a new category of insurance tailored to online scenarios in North America, powered by its proprietary AI insurance engine. By combining deep learning–based actuarial models with API distribution, Seel customizes big data–driven insurance products for various industries.

Unlike traditional insurers, which rely on actuaries to calculate prices based on historical data and comparable market rates, Seel’s products feature high-frequency and low-premium traits—making them naturally suitable for machine learning. Instead of traditional actuarial tables, Seel uses neural networks to reimagine the production and distribution of insurance for high-frequency, low-value use cases.

Traditional insurers acquire customers mainly through agents, incurring high customer acquisition costs. For example, acquiring a customer for a $1,000–$1,400 auto insurance policy in the U.S. typically costs around $200–$300. Seel’s products, priced between $1 and $10, are incompatible with such expensive acquisition models. Instead, Seel focuses on a B2B approach, embedding its products into real-life usage scenarios via APIs.

Founded in November 2019, Seel initially concentrated on building its API distribution engine and risk assessment (pricing) system. In 2021, it quickly identified its first product-market fit: Refund Insurance, which officially launched in August that year.

Refund Insurance differs from Shipping Insurance. While Shipping Insurance covers return shipping costs, Refund Insurance covers the value of the product itself.

The U.S. e-commerce market is projected to reach $1 trillion by 2022, and returns are a major pain point. Unlike China, where platforms like Alibaba and JD dominate and independent brands are scarce, the U.S. e-commerce market is largely composed of independent brands. Even Amazon, the largest e-commerce player, only accounts for about a third of total GMV.

In the U.S., most e-commerce platforms offer a 30-day no-questions-asked return policy. Including shipping time, it takes nearly two months for merchants to confirm their net revenue. During this time, merchants must bear cash flow risks or seek financing to bridge the gap.

To mitigate this, Seel introduced two Refund Insurance products. The merchant version uses a proprietary big data model to analyze customer behavior, product details, and more to calculate the likelihood of a return and corresponding premium at the moment of purchase. If the customer returns the product, Seel covers the refund amount, protecting the merchant’s cash flow. Returned items can be resold or donated for tax credit purposes. In this way, brands can pay a premium to offset cash flow uncertainty and immediately recognize revenue.

Seel also launched a consumer-facing Refund Insurance tailored to non-refundable items. For instance, merchants often sell heavily discounted items on Black Friday with no return policy. Seel allows these merchants to embed Refund Insurance into their websites, giving customers the option to add protection at checkout. If unsatisfied, the buyer can return the product and receive a refund through Seel.

For platforms with strict return policies (e.g., 7 or 14 days), Seel also offers Return Window Extensions. Buyers can purchase Refund Insurance to extend the return period to 30 days.

This product performed beyond expectations, with an attach rate of 20–25%—about 1 in 4 or 5 consumers opted to buy it at checkout.

Simply put, Seel eliminates refund concerns for both merchants and consumers. Merchants increase sales and improve cash flow certainty, while consumers gain peace of mind when purchasing items they might otherwise hesitate to buy.

How does Seel manage its own risk? The answer lies in pricing—and for an AI insurance company, pricing is all about the algorithm. Unlike traditional actuarial models, Seel’s deep learning approach can process unstructured data such as images, timestamps, sequences, and text. For Refund Insurance, this includes product images (via computer vision), product descriptions (natural language processing), and purchasing behaviors (via LSTM networks), generating highly detailed user profiles. These data types are out of reach for traditional actuarial tables, giving Seel a data advantage.

Deep learning models discover many more correlations among variables than human actuaries. “We built a refund model with credentialed actuaries that identified around 30 relevant factors,” said CEO Peng Cheng. “But our deep learning model found over 2,000 in just one day—it’s not even close.”

As Seel grows and signs more merchants, it gains access to more data, which improves its models and predictive accuracy. This positive feedback loop creates a powerful network effect, eventually forming a technological moat. New competitors struggle to catch up without similar data volume. This actuarial network effect is similar to Google’s search engine or Netflix’s recommendation engine—the more users, the better the product, and the more attractive it becomes.

Refund Insurance isn’t entirely new. Companies like Stripe and Riskified also offer services that let merchants pay premiums to protect against chargebacks—a risk that represents only about 1% of sales. But return rates typically range from 10% to 30%, which makes Seel’s solution even more valuable.

As a result, Refund Insurance was well received in North America. In just four months of beta testing, and with zero marketing spend, Seel signed hundreds of independent e-commerce brands, including Garmentory, one of the largest fashion platforms in the U.S.

Beyond Refund Insurance, Seel’s actuarial engine has applications in other data-rich industries. The company is now developing two new products: On-Time Delivery Insurance and Income Insurance. Seel has already signed agreements with food delivery and logistics platforms for these upcoming offerings.

“We want to build a whole new category of insurance around high-frequency, low-premium products,” said CEO Peng Cheng. “Refund Insurance is just the beginning. We’re also developing On-Time Delivery and Income Insurance products for gig workers like couriers and drivers. These will launch later this year.”

Seel’s core team hails from deep-learning companies like Orbital Insight and Google. CEO Peng Cheng was a lead engineer at Orbital Insight and holds dual bachelor’s degrees in philosophy and physics from UNC Chapel Hill. COO Zhang Sinong also graduated in philosophy from UNC and co-founded Seel. CTO Chen Mo built big data systems at Orbital Insight and previously worked at Google on the Next Billion Users project, focusing on data and privacy tools.

After initially investing in Seel in late 2019, Mee-Jung Jang joined full-time as CRO in early 2021. She was previously a partner at Techstars Ventures and holds a dual BA in computer science and economics from Harvard and a JD from Yale Law School.

Seel’s primary market is the United States. Following strong merchant growth in just four months of product launch, the company plans to expand its insurance offerings and open a new R&D hub in Shanghai this year.

Throughout financial history, new categories of insurance often emerge alongside new technologies. Just as Edmond Halley’s life table enabled life insurance and Ford’s standardized Model T unlocked auto insurance, Seel is leveraging big data to create a new class of insurance. Its expansion across sectors and geographies is well worth watching.

‍