AI Startup Slope Raises $30 Million, Launches B2B Payment Tracking and Collection Platform

Tracking and ensuring payment integrity is not an easy task for any business, but in the B2B field, there are additional difficulties. It is necessary to ensure that the companies you sell to will actually make payments on time, without unjustified refusals, bankruptcies, or other ways of evading payment. These issues are much less common in the B2C sector.

B2B companies, such as wholesalers selling large quantities of products, face the challenge of ensuring that they are paid for the products and services they provide on time. San Francisco-based AI startup Slope, founded two years ago, is attempting to create an industry standard: a B2B payment tracking and collection platform, partially driven by its own rules-based technology and partially powered by OpenAI's GPT-3.5 Turbo. Slope is also developing its own proprietary Large Language Models (LLMs).

The company announced today that it has completed a $30 million equity financing round, led by renowned venture capital firm Union Square Ventures' Fred Wilson, with participation from OpenAI's CEO and co-founder Sam Altman, bringing the total raised to $187 million. This is a significant amount for a lean team of only 18 full-time employees.

Slope's CEO and co-founder Lawrence Lin Murata said in a video call with VentureBeat, "Our operations are highly efficient." Lin Murata said he personally experienced the challenges faced by B2B suppliers in the wholesale goods business they operated in their home country of Brazil.

He added, "For B2B companies, being able to quickly obtain funds and send them to their suppliers is vital."

Slope's technology covers the entire B2B customer payment process

The company will use this funding to continue building its team and technology, including an online payment and invoicing tool that Slope's customers can use to accept payments from their own customers, including credit cards, ACH (Automated Clearing House), and international payments.

Slope co-founder and Chief Product Officer Alice Deng said, "We handle everything from customer registration and risk assessment to reconciliation and everything in between." She also stated that Slope analyzes buyer's "credit risk, invoicing, billing, cash applications, deductions, and more, all the way to synchronizing with the B2B customer's accounting system, all on our platform."

Slope also provides financing for its customers' customers, allowing those who are unable to make upfront payments to obtain credit directly through Slope's payment system.

Deng also stated that Slope offers a new "visibility" to the B2B payment process, as these processes can be "outdated" and less transparent.

One example of this new visibility is Slope Timeline, a feature that leverages Slope's understanding of enterprise transactions to provide near real-time information on payment and product delivery status to B2B suppliers and their customers.

Deng said, "Both the buyer and the seller will have a precise understanding of their stages, down to the millisecond." This is a significant improvement compared to traditional enterprises often wondering, "Has my order been initiated? Has it been shipped? Has my wire transfer been reconciled?"

The foundation of "clean data"

The key to Slope's approach lies in obtaining "clean data" from enterprise customers to understand and provide up-to-date information about B2B payments and help mitigate risks.

Deng said, "This is indeed the foundation of our system, driving everything we do." "We are an AI company, but in reality, we are a clean data company."

To obtain "clean data," Slope collaborates with enterprise customers to collect all data related to the orders they receive, process, and ship.

"This data is formatted and presented on our platform in a way that is useful to our customers," Lin Murata said.

SlopeGPT leverages the advantages of GPT in a novel way, transforming enterprise transaction data into a fraud risk analysis tool

This is where some of Slope's AI methods come into play: Slope is able to assess the creditworthiness and fraud risk of B2B buyers and utilize SlopeGPT to provide optimal credit to sellers/Slope's customers, reducing risks. This tool was launched in April.

SlopeGPT uses transaction and purchase order data from Slope's enterprise customers and clusters the data using a dedicated OpenAI GPT instance (not accessible on the public internet or by third parties) to determine which types of payments are normal and which are abnormal. Slope then uses these clusters and other rule-based data management techniques to provide relevant data and recommendations to its customers and their customers.

In the process of analyzing buyer's risk to grant them financing/credit, SlopeGPT can also look for clues.

Lin Murata said, "If there is abnormal activity, or if someone pretends to be another company, or steals another company's information... if they deliberately try to show a strong cash flow and then suddenly perform fund transfers, we can detect these anomalies to prevent payment."

Slope discovered the power of GPT by inputting 2.5 million bank transactions it conducted over 18 months into the GPT.

The founder stated that the company has also developed its own proprietary LLM, which performs better in accurately identifying risks after training on public data, and will be released soon.