Ecommerce / B2B Payments · Deep dive

Slope

AI-underwritten B2B net terms and financing embedded into merchant checkout and invoicing — $77M equity plus $175M debt raised, JP Morgan Payments as lead investor and lender, now powering credit for Amazon sellers and Fiserv's SnapPay.

emerging

The question that decides it: Does Slope's AI underwriting plus embedded-checkout wedge into B2B commerce hold up as credit losses normalise post the 2023-2024 rate cycle, and does merchant-integration depth (Fiserv SnapPay, Amazon sellers, JP Morgan Payments distribution) create switching cost — or do Stripe and Adyen ship comparable underwritten B2B net terms inside the rails merchants already use, collapsing standalone B2B BNPL into a feature line the way Klarna and Affirm compressed independent consumer BNPL? Answer conditions: (a) sustained loss rates in the 1-2% range across a full rate cycle, published or leaked to lenders; (b) at least one more anchor distribution deal at Amazon or Fiserv scale by end-2027 that Stripe or Adyen were shopped and lost; (c) the JP Morgan Payments relationship deepens into exclusive product distribution rather than a warehouse line; (d) revenue take-rate remains above 2% blended without meaningful attrition to a payments incumbent's bundled offer. Fail two of the four and Slope becomes a good acquisition for JP Morgan Payments — not an independent outcome.

My take ★★ Very interesting ★ Interesting − Maybe ✗ Not interesting

Snapshot

Slope sells enterprise merchants a B2B checkout and invoicing product that lets their business buyers pay on net-30/60/90 terms or in installments, with Slope fronting the cash and taking the credit risk. Founded in San Francisco in 2021 by two second-time founders — Lawrence Lin Murata (ex-Nauto) and Alice Deng (ex-Dropbox, ex-Glisten YC W20) — the company has raised roughly $77M equity across a YC S21 seed, 2022 Series A, September 2023 USV/Sam Altman extension and July 2024 strategic round led by J.P. Morgan Payments, plus ~$175M of debt facility, for a reported $252M total. As of Dec 2025, Slope powers Amazon’s seller working-capital program (JPM-backed); as of 2025 it sits inside Fiserv’s SnapPay AR product. Headcount is ~38 (Tracxn, Mar 31, 2026) — either an efficiency story or a warning.

Founding story

Murata’s route into fintech ran through autonomous vehicles. Newton Technologies, his first company (Stanford CS), used smartphone sensor data and ML to score unsafe driving; Nauto acquired it in February 2019, and Murata ran AI Platforms and Data Science there. The Newton engineers became the initial core he brought to Slope two years later. Deng came via consumer product — UC Berkeley, then Dropbox Paper PM, then Glisten.ai (YC W20). She and Murata met in 2016 running rival Berkeley and Stanford hackathons and reconnected in 2021 to start Slope.

The Sam Altman connection traces to Murata sitting in on Altman’s 2016 Stanford ‘How to Start a Startup’ class; it resurfaced in Sep 2023 when Altman put a personal check into the Series A extension, and in Jul 2024 when Jack and Max Altman (via Saga) came in. Getting Altman family capital twice in ten months is unusual; getting J.P. Morgan Payments as both lead and warehouse lender in the same transaction is rarer. The founding grievance is standard — B2B buyers want net terms, forcing sellers into an ad hoc credit shop with manual underwriting and DSO drag — and Slope’s answer was to build the underwriting, checkout and balance sheet inside one product.

How it works

Three things happen per transaction. The merchant embeds Slope into its checkout, invoicing portal or ERP flow (via API, and increasingly via Fiserv’s SnapPay). When a business buyer picks a term, Slope runs KYB and underwriting — instant approval for smaller lines, within two business days for larger facilities. Slope then pays the merchant upfront out of its own capital (drawn from the J.P. Morgan warehouse and other debt), less its fee. Slope services collections at term and takes the credit loss on default.

The underwriting stack is the load-bearing part. Slope markets two AI systems by name: SlopeGPT, a transformer-based model that ingests business bank-transaction data (typically Plaid-fed) and clusters cash-flow patterns to score risk — described on the company’s Medium as ‘the first payments risk model powered by GPT’ (April 2023); and Slope TransFormer, trained to parse bank-statement language. These run alongside bureau data and internal loss history. The claim that SlopeGPT catches risks conventional models miss is not independently audited, but J.P. Morgan’s July 2024 decision to both lead the equity round and extend a warehouse is a form of validation — a warehouse lender diligences the underwriting.

Approval limits per company disclosure run up to $250,000 instantly for smaller buyers; on the December 2025 Amazon program, top-tier sellers access revolving credit up to $5M with 2-12 month terms per draw, and Bloomberg reported Slope typically charges ~2.5% on a 60-day loan there (roughly 15% APR) — which frames the unit economics.

Product and business overview

Three product surfaces sit on the same rails. Slope Pay is the merchant checkout: net 30/60/90, instalments, ACH, wire and card, plugged into a seller’s web or invoicing flow. SlopeAI is the underwriting and bank-data enrichment platform (SlopeGPT, TransFormer, KYB/KYC agents), now sold to financial institutions and wholesalers that want the models without the white-label checkout — a separate revenue line. Slope credit is the working-capital line, most visibly the December 2025 Amazon program where eligible U.S. sellers access pre-approved revolving credit up to $5M on J.P. Morgan-backed capital.

Named distribution partnerships as of Aug 2026: Fiserv SnapPay (reaching JD Edwards, Microsoft Dynamics, Oracle and SAP ERPs); J.P. Morgan Payments (embedded net terms for JPM’s enterprise clients); Amazon (seller working capital, Dec 2025); LabConnect (clinical-trials receivables, expanded Q4 2025); Plaid (bank-data infrastructure). The direct merchant book is not publicly enumerated.

Business model and pricing

Slope monetises three ways. Primary is a merchant transaction fee on financed volume — from as low as 1.60% for Net 30 on the pricing page, bespoke per lane and buyer risk. Second is interest on credit lines — APRs from 9.00% per the site, with the Amazon program at ~2.5% on a 60-day draw (~15% APR) per Bloomberg. Third is SaaS-style fees for the order-to-cash platform, per Sacra.

A 1.60% Net 30 rate implies ~19-20% APR before losses on the financed cash, out of which Slope pays cost of capital (whatever JPM charges), underwriting losses, CAC and infrastructure. Whether the fee passes to the merchant (widens funnel, narrows margin) or the buyer (opposite) matters more than the sticker level. The company’s FAQ language — ‘Slope’s merchant partners work with Slope to determine fees, and if any of the cost will be passed onto a Buyer’ — captures how much the economics depend on the counterparty, and is a large part of why no audited loss rate is public.

Traction over time

No revenue, GMV or loss-rate figure has been published as of Aug 2026 — the most important gap in any read on the company.

Market analysis

Slope’s internal TAM number is the $125T global B2B payments flow (TechCrunch, Sep 2023) — the volume across which a B2B payments company hopes to skim basis points, not the wallet. Sacra frames the relevant slice as accounts-receivable automation at $2.8B (2024) growing to $6.4B by 2033. In Europe, GlobeNewswire (Apr 2026) puts the B2B BNPL segment at $80.2B GMV in 2025, projected to $179.6B by 2030. B2B trade credit is the world’s largest source of working capital, historically underwritten by trade-credit insurers and factored expensively; moving it online creates room for a technology take.

The counter-force is that consumer BNPL, the more mature analogue, is visibly stress-testing. The CFPB’s Dec 2025 report and Richmond Fed’s 2026 brief both flagged rising delinquency; Lending Tree reported 42% of BNPL users made at least one late payment in 2025, up from 34% in 2023. B2B losses are typically more stable than consumer, but B2B BNPL is a young category with limited cycle-tested loss data and U.S. small-business bankruptcies rose sharply through 2024. Hokodo’s late-2025 wind-down in Europe is the first concrete signal the category can fail.

Competitive intel

Three rings. Direct B2B BNPL peers: Resolve is the U.S. comparable (Affirm spinout, 2019, $60M raised, SMB-heavy, more transparent pricing); Balance is the checkout-first analogue with Stripe, Ribbit and Max Levchin on the cap table. In Europe, Mondu ($70M), Billie ($150M+, $640M peak Oct 2021), Two and Kriya lead — same architecture. Hokodo winding down in late 2025 is the first published category casualty.

Enterprise incumbents: TreviPay (formerly MSTS) — managed B2B trade credit for global brands, ~$7B annual volume, Corsair-owned. The market Slope is climbing into from the checkout end. C2FO sits adjacent with a different mechanic (dynamic discounting, $100M+ revenue, SoftBank/Temasek).

Payments rails: Stripe, Adyen and, via the Amazon partnership, Amazon itself. None ship underwritten B2B net terms at Slope’s depth as of Aug 2026, but all have distribution Slope does not. The consumer-BNPL lesson: when Apple Pay Later, Shop Pay Installments and PayPal Pay in 4 shipped inside the rails, the standalone brands lost pricing power. Vartana’s June 2025 acquisition by Capchase is the closest recent M&A comp and the most plausible template for how Slope resolves if the standalone path narrows.

History and evolution

The dated table already covers the raises; three inflection points bear reading behind it. April 2023’s SlopeGPT launch — narrative or real underwriting lift, it teed up the September Altman-led extension. July 2024’s JPM Payments round: a systemically important bank leading equity and providing a warehouse in the same transaction is not casual; the read was that JPM intended to distribute Slope to its own enterprise base, which is what happened via Fiserv in 2025 and Amazon in December 2025. The Amazon program effectively re-sells Slope’s underwriting stack under Amazon’s brand using JPM’s capital. The 38-person March 2026 headcount says the founders are choosing operating leverage from AI plus strategic distribution over a scaled sales army. No public stumbles — but no revenue, GMV or loss-rate figure either, which is itself a data point.

What people say

The case for. Merchant reviews on SourceForge, ProductHunt and FeaturedCustomers call out the checkout integration, buyer-approval speed and the API. Plaid’s customer story frames Slope as a reference deployment for cash-flow underwriting on bank-transaction data. The sales motion is sticky — once Slope is live inside a merchant’s checkout or ERP, switching cost is real. J.P. Morgan Payments’ July 2024 decision to both lead equity and provide the warehouse is the strongest single signal in the file.

The complaints. The recurring negative theme in reviews (Resolve’s comparison pages, SourceForge) is customer support — competitors cited as more responsive, unsurprising for a 38-person team supporting large enterprise integrations. Structurally, no independently audited loss rate is public. B2B BNPL is credit lending dressed as software; if losses exceed assumptions, the P&L unwinds through the warehouse. Hokodo’s late-2025 wind-down is the European reference case; Vartana-into-Capchase (June 2025) is the M&A one — the outcome shape for B2B BNPL brands that miss escape velocity is absorption, not IPO.

Outlook: the open question

What has to be true for Slope to be an independent outcome, not an acquisition, is a specific list, and the market is running the test now. Slope took the SlopeGPT AI-underwriting narrative, converted it into a J.P. Morgan Payments strategic round plus warehouse in July 2024, and translated that capital into anchor distribution — Fiserv’s SnapPay in 2025, Amazon sellers in December 2025. That is close to the maximum extractable from the founder-plus-narrative-plus-strategic-lender playbook in three years.

The failure modes are specific. Credit: no loss rate is public; B2B BNPL is early cycle; Hokodo wound down in late 2025; if 2025-2026 loss experience runs above assumptions, the JPM warehouse reprices or pulls and equity is impaired. Distribution: Stripe and Adyen have not shipped comparable underwritten B2B net terms as of Aug 2026, but they own the checkouts Slope wishes it did — Apple Pay Later, Shop Pay Installments and PayPal Pay in 4 shipping inside the rails is the compressor template. Strategic investor: a lead that also lends your book has the optionality to turn a partnership into an acquisition on unfavourable terms if the company stumbles.

Answer conditions: sustained 1-2% loss rates published to lenders through a full cycle; another Amazon- or Fiserv-scale distribution deal by end-2027 that Stripe or Adyen were shopped and lost; the JPM relationship deepening into exclusive product distribution rather than just a warehouse; blended take-rate holding above 2% without material attrition to a bundled payments-incumbent offer. Fail two of four and Slope is a very good acquisition for JPM Payments, not an independent business. Bull case: Slope becomes the underwriting layer inside every enterprise B2B checkout, the way Plaid became the connectivity layer inside every fintech account link. Bear case: Vartana-into-Capchase at ten times the size.

How a challenger would attack it

Attack the opacity and the headcount. Slope publishes no revenue, no GMV, no loss rate, and its pricing is bespoke — the FAQ’s “merchant partners work with Slope to determine fees” language is a negotiation black box. A challenger would do what Resolve already half-does: publish transparent rate cards, publish audited loss performance, and turn Slope’s silence into the sales objection. Second: service. The recurring review complaint is unresponsive support, the predictable output of 38 people carrying enterprise integrations at Fiserv and Amazon scale; a challenger staffing real implementation and support teams wins the merchants who found Slope’s API elegant and its humans absent. Third, and structural: Slope’s distribution is rented. Fiserv, Amazon and JP Morgan own the channels; the Amazon program re-sells Slope’s stack under Amazon’s brand on JPM’s capital, meaning the challenger’s cleanest move is to be the payments rail itself — a Stripe or Adyen shipping underwritten net terms inside checkouts merchants already run, the exact compression that Apple Pay Later and Shop Pay Installments executed on consumer BNPL. Underwriting on bank-transaction data is not a moat when Plaid sells the same feed to everyone; the moat was distribution, and Slope’s belongs to its partners.

Same playbook, new buyer

The playbook — AI cash-flow underwriting plus embedded terms, funded by a bank warehouse — ports to buyers Slope’s enterprise motion skips. The obvious first shift is down-market: Resolve holds SMB, but the mid-market wholesale and distribution tier — too big for instant $250K lines, too small for TreviPay’s managed programs — has no default provider, and a self-serve product with published pricing would take it. Second: vertical B2B marketplaces. Slope’s LabConnect deal (clinical-trials receivables) is the prototype it hasn’t generalized — construction materials, food service distribution, medical supplies and auto parts each have marketplace consolidators that need a captive credit layer, and a vertical-specialist underwriter beats a horizontal one on loss data within its niche. Third: geography — Hokodo’s wind-down left European B2B BNPL ($80.2B GMV in 2025, headed to $179.6B by 2030) to Mondu and Billie with no US-grade AI-underwriting entrant. Slope can’t chase any of these: 38 people, a JPM warehouse denominated in US enterprise risk, and anchor partners (Amazon, Fiserv) who consume all available integration capacity. The SlopeAI models-only line even shows the company knows its underwriting travels further than its team can.

Sources and further reading

• How a B2B payments startup won Max, Jack and Sam Altman, JP Morgan as investors — TechCrunch, July 17, 2024
• Sam Altman backs Slope’s $30M round to digitize, scale B2B payments — TechCrunch, September 27, 2023
• Slope, AI-Led B2B Payments Platform, Secures $65 Million — BusinessWire, July 17, 2024
• Amazon Partners With JPMorgan-Backed Slope to Offer Seller Loans — Bloomberg, December 16, 2025
• Slope Working with Amazon to Expand Access to Capital for Amazon Sellers — Slope, December 16, 2025
• SlopeGPT: The first payments risk model powered by GPT — Slope Stories on Medium, April 2023
• Slope funding, news & analysis — Sacra, updated 2025-2026
• Slope: The B2B Payments Platform — Y Combinator company page
• Slope - Company Profile, Team, Funding & Competitors — Tracxn, 2026
• BNPL Market Report 2025-12 — CFPB, December 2025
• Buy Now, Pay Later: Recent Developments and Implications — Richmond Fed, 2026
• Europe B2B Buy Now Pay Later Business and Investment Report 2026 — GlobeNewswire / Research and Markets, April 30, 2026