Affirm Interview Guide 2026: BNPL, Underwriting, Levchin Culture, Post-CFPB Engineering

Affirm Interview Guide 2026: Buy-Now-Pay-Later, Underwriting, Merchant Network, and the Levchin Engineering Culture

Affirm (NASDAQ: AFRM) is the largest US-headquartered buy-now-pay-later (BNPL) company. Founded in 2012 by Max Levchin (PayPal co-founder), the company built its position on transparent installment loans — fixed terms, no late fees, real underwriting on each transaction — differentiating from credit-card-style revolving debt. The hiring process reflects Levchin-era engineering values: technical depth, written reasoning, conviction about the product mission. The 2024–2025 BNPL regulatory shift in the US (Consumer Financial Protection Bureau treating BNPL as credit-card-equivalent under TILA) has reshaped product engineering. This guide covers what Affirm does, the engineering tracks, the interview process, and what makes Affirm hiring distinctive in 2026.

What Affirm Does

Affirm operates a consumer BNPL platform plus a card product:

Pay in 4: short-term installment plans (4 payments over 6 weeks), interest-free.
Monthly installment loans: longer-term loans (3, 6, 12, 24, 36 months), interest-bearing for some merchants and consumers, 0% APR for others.
Affirm Card: debit-card-with-installment-options consumer product launched 2022 — debit by default, can elect installment financing on individual transactions.
Merchant network: integrations with major retailers (Amazon partnership renewed 2024, Walmart, Target, Best Buy, Apple, Shopify, etc.).
Adaptive Checkout: the merchant-facing product that prices and offers loans dynamically at checkout.
Affirm Money Account: savings account product launched alongside the Card.
Affirm SDKs: Shopify integration, native iOS / Android SDKs, web SDK for merchant integration.

Distinctive features:

Per-transaction underwriting: unlike credit cards which extend a revolving line, Affirm underwrites each transaction. The underwriting models are central to the business and a major engineering investment area.
Transparent / no-late-fee positioning: Affirm explicitly avoids late fees and revolving debt — the marketing distinction that drives consumer adoption.
Levchin-era engineering culture: direct, written-reasoning-heavy, technically deep. The “Affirm Reading List” and intellectual culture are real.
Public company: NASDAQ: AFRM since 2021; substantial scrutiny.
Regulatory exposure: the 2024–2025 CFPB rule treating BNPL as credit-card-equivalent under TILA has reshaped compliance engineering. The Trump administration’s CFPB transition adds further regulatory uncertainty.

Roles Affirm Hires For

Software engineer (backend / loans platform)

Builds the loans platform — application processing, underwriting integration, servicing, payments, collections. Python primary; Go in some newer services. Distributed systems with strict consistency requirements (financial transactions).

Software engineer (merchant / integrations)

Builds merchant SDKs, checkout flows, partner integrations. JavaScript / TypeScript heavy for web SDK; mobile SDKs in Swift / Kotlin.

Software engineer (Card / consumer)

Affirm Card platform — card issuing (Stripe Issuing or in-house), transaction routing, installment-option-on-debit logic. Hybrid of fintech and consumer-product engineering.

ML engineer / data scientist (underwriting / risk)

Underwriting models — credit decisioning, fraud detection, repayment prediction. Substantial investment area. Models drive pricing and approval; engineering quality directly affects business outcomes.

Data engineer

Pipeline engineering, feature platform for ML models, regulatory reporting infrastructure. Substantial scale (transactions across millions of consumers).

Frontend engineer

Affirm consumer web and app, merchant dashboards, internal admin tools. React + TypeScript dominant.

Compliance / regulatory engineer

Engineering work for compliance (truth-in-lending disclosures, state-level licensing, CFPB regulatory reporting). Growth area post-2024 regulatory shift.

Security / fraud engineer

Fraud detection, account security, transaction security. Adversarial engineering with continuous attacker pressure on identity / synthetic-identity fraud.

Affirm Interview Process

Round 1: Recruiter screen

30 minutes. Background, motivation, role fit. Affirm recruiters often probe specifically on product engagement — whether you understand BNPL and have considered the trade-offs.

Round 2: Technical phone screen

60–90 minutes. Coding (medium-hard), some technical depth. Practical engineering flavor; less algorithmic gymnastics than FAANG.

Round 3: On-site / virtual on-site

4–5 rounds, each 60–90 minutes:

Coding (1–2 rounds) — practical engineering, often with fintech / financial-systems flavor
System design (1 round) — fintech systems (loan servicing, payments, idempotency, ledger consistency)
Domain depth (1 round) — depends on role: ML for underwriting, mobile, distributed systems, compliance engineering
Behavioral / values fit (1 round) — collaboration, written reasoning, conviction about the product mission

Round 4: Decision

Calibration meeting; offer typically within 1–2 weeks. Compensation negotiation expected.

What Affirm Tests For

Practical engineering and financial systems thinking

Affirm coding rounds emphasize practical fluency — financial system correctness, edge case handling, idempotency, audit trails. Engineers from FAANG-prep tracks sometimes over-rotate on LeetCode and miss the financial-systems-judgment signal.

Written communication

Levchin-era culture rewards written reasoning. Expect interview questions that probe written communication — how you’d document a design decision, how you’d write a memo about a trade-off. Some interviews include a written component.

Conviction about the product mission

Affirm’s positioning (transparent loans, no late fees) is mission-driven. The behavioral round probes whether you genuinely believe in the product or are just there for the comp. Engineers who treat it as mission-fit screening underperform if they’re indifferent.

Regulatory awareness (post-2024)

The 2024 CFPB rule and the broader BNPL regulatory environment matter. Engineers expected to engage with compliance dimensions; “I just write code, regulation is someone else’s job” attitudes don’t fit Affirm culture.

ML pragmatism (for ML roles)

Underwriting ML has real-world consequences (approvals affect consumers, model errors cost money). Strong ML candidates demonstrate they think about model lifecycle, monitoring, fairness, and adversarial pressure — not just model architecture.

Compensation

Competitive, with substantial RSU component:

New-grad SWE: $170k–$260k total comp first year
Mid-level (4–7 years): $250k–$400k
Senior (8+ years): $380k–$600k
Staff / Principal: $550k–$1M+

Compensation is RSU-heavy. AFRM stock has been volatile — substantial decline 2021–2022 (from $175 IPO highs to ~$10 lows), recovery through 2023–2025 (now in the $30–$60 range depending on quarter). Engineers joining at low stock prices saw substantial appreciation; current entries face uncertainty.

Working at Affirm

Tech stack and engineering quality

Python dominant for backend; React + TypeScript for frontend; Swift / Kotlin for mobile; Go in newer services. Engineering quality is regarded as high; the financial-systems discipline (testing, audit trails, idempotency) is real.

Pace and intensity

Moderate. Less frenetic than ByteDance-style tech; more intense than mature FAANG. Fintech regulatory deadlines drive some intensity; team cultures vary.

Office and remote

HQ in San Francisco. Major offices in NYC, Toronto, Salt Lake City, Chicago. Remote-first since 2021; some hybrid expectations for senior roles.

Career trajectory

Standard tech-style leveling. Promotion is rigorous; the bar for senior+ levels is real. Strong performers see steady progression; the engineering culture rewards depth over volume.

Affirm vs Alternatives

Affirm vs Klarna: The largest BNPL competitor globally. Klarna is European with US presence; Affirm is US-focused. Klarna more consumer-app-flavored; Affirm more transaction-financing-flavored. Klarna IPO’d in 2025 after years of delays.

Affirm vs Afterpay (Block): Afterpay was acquired by Block in 2022; now operates as part of Block’s Cash App ecosystem. Different product positioning post-acquisition; consumer-app integration vs Affirm’s standalone positioning.

Affirm vs PayPal Pay Later: PayPal added BNPL in 2020. Affirm has deeper underwriting; PayPal has larger consumer base and wallet integration. Different positioning; both compete for merchants.

Affirm vs traditional fintech (Stripe / Square): Different product surface. Stripe is broader payments infrastructure; Square is small-business commerce; Affirm is consumer-credit-at-checkout. Engineering work different.

Things That Surprise Candidates

The Levchin-era culture is real — written reasoning, technical depth, conviction matter more than at most peer fintechs.
The regulatory exposure is substantial; compliance engineering is a major focus area post-2024 CFPB rule.
The underwriting ML is technically deep; ML engineers from web / consumer backgrounds have a learning curve.
Stock volatility has been substantial; engineers should calibrate equity expectations carefully.
The mission framing is genuine; engineers who treat it as marketing don’t last.

Frequently Asked Questions

How does the 2024 CFPB rule affect Affirm engineering?

Substantially. The rule treats BNPL as credit-card-equivalent under TILA, requiring disclosures, dispute resolution processes, and refund mechanisms similar to credit cards. Engineering work to build TILA-compliant flows, dispute systems, and statement formats has been substantial. The regulatory landscape continues evolving under the Trump-administration CFPB; engineers track these closely.

What’s the underwriting ML actually like?

Technically deep and high-stakes. Models combine consumer credit data, transaction context, merchant signals, and behavioral features. Approval decisions and pricing are model outputs; errors directly affect business performance. Engineers work on model architecture, feature engineering, monitoring, fairness analysis. The work is more like quant trading model engineering than typical consumer ML.

How is Levchin’s culture really expressed in engineering?

Written reasoning, technical depth, and intellectual conviction matter. Internal documents are heavy; design decisions are written and debated. Engineers describe more written communication than at most peer companies. Levchin himself has been visible in engineering culture, including via the Affirm Reading List and direct technical engagement on hard problems.

Is Affirm a good place for early-career engineers?

Yes for engineers interested in fintech and rigorous engineering culture. Mentorship is generally good; the engineering depth is real. New-grads ramp into specialty teams (loans platform, underwriting ML, merchant integrations, etc.) and develop deep expertise. The financial-systems discipline transfers well to other fintech roles.

How risky is the BNPL business given regulatory pressure?

Real but managed. Affirm has been more compliance-engaged than some BNPL peers; the per-transaction-underwriting model aligns better with traditional credit regulation than revolving-debt models. Ongoing regulatory evolution adds uncertainty; engineers should expect compliance to remain a major engineering theme.