Aakash GuptaHow to Build AI Products in FinTech ($100B Robinhood VP Lessons)
CHAPTERS
- 0:00 – 1:34
Robinhood at $100B: Meet the VP behind product velocity
Aakash frames Robinhood’s rapid market-cap rise and introduces Abhishek Fatehpuria, VP of Product. The episode sets expectations: practical lessons on building high-trust FinTech products, including AI, IPO access, and team scaling.
- •Robinhood’s recent stock and market-cap growth as context
- •Introducing Abhishek Fatehpuria’s role and scope
- •Tease of key themes: AI in FinTech, democratizing access, product craft
- 1:34 – 3:03
Cortex AI Assistant: Stock Digest, workflows, and feature roadmap
Abhishek explains Robinhood Cortex as an investing assistant that fits into what users already do—especially answering the immediate question after a price move: why did it move? He shares the initial Cortex capabilities and where it may expand next (e.g., crypto).
- •Stock Digest: explain price movement after notifications
- •Data sources combined: news, research, Robinhood trading activity, SEC filings
- •Workflow-first AI: solve existing user pain vs. “AI for AI’s sake”
- •Roadmap: roll-out for stocks first; crypto soon
- 3:03 – 5:53
AI in a regulated brokerage: data curation, prompting, and trust guardrails
The conversation turns to the constraints of AI in FinTech: regulation, user trust, and avoiding harmful outputs. Abhishek details how Robinhood tightly curates input data and designs Cortex to be informational rather than advisory—for now.
- •Regulatory constraints and money-handling raise the trust bar
- •Curated inputs and licensed sources as a foundational safeguard
- •Prompting/coaching the model to avoid mistakes and recommendations
- •Sequencing: information tools first; advisory/recommendations later
- 5:53 – 8:01
Building Cortex the Robinhood way: problem selection and product discovery
Abhishek describes their development approach: identify high-frequency customer problems, then apply AI where it clearly improves outcomes. He outlines the two early Cortex demos—digest and trade builder—and why these were prioritized.
- •Start from known user problems, then apply AI
- •Two initial features: Stock Digest and Trade Builder
- •Trade Builder helps translate hypotheses into stock/options trades
- •Internal utility: teams also consult Cortex for context
- 8:01 – 9:37
General advice for FinTech AI PMs: regulation literacy, feasibility, and patience
Abhishek offers broader guidance for PMs building AI in financial services. Success requires understanding what the tech can do, what regulators will allow, and rolling out incrementally so customers and compliance teams build confidence.
- •Learn the technology’s realistic capabilities
- •Become fluent in the regulatory landscape
- •Start small and iterate; multi-year vision is fine
- •Trust is earned gradually—especially with money
- 9:37 – 12:38
Tokenization in Cannes: stock tokens and private-company access demand
Aakash asks about Vlad’s Cannes presentation, and Abhishek explains the tokenization announcements. The discussion focuses on global access to US equities and the strong retail desire to invest in private companies staying private longer.
- •Stock tokens: stablecoin-like rails for US stocks abroad
- •Motivation: global demand for US equities (Apple, Tesla, Nvidia, etc.)
- •Private stock tokenization taps unmet retail demand
- •Why it resonated: retail wants access to high-value private companies
- 12:38 – 16:47
IPO Access: how retail gets shares at the IPO price
Abhishek tells the origin story of IPO Access and explains the mechanics: Robinhood joins the selling group, collects retail indications of interest, shares demand with underwriters, then allocates shares the night before trading. He contrasts this with legacy brokers that reserve IPOs for high-net-worth clients.
- •Launched May 2021; started late 2020 around Robinhood’s own IPO thinking
- •Mechanics: selling group role, demand collection, allocation process
- •Allocation constraints: retail demand often exceeds shares received
- •Democratization angle: IPO participation previously gated to high-net-worth users
- 16:47 – 25:10
Robinhood’s innovation DNA: value + delightful experience, “swipeys,” and pixel craft
Aakash probes how Robinhood consistently ships innovative products without getting bogged down by compliance friction. Abhishek outlines their product recipe (high customer value plus delight), the constraint-driven innovation of FinTech, and a key working-backwards tool: drafting “swipeys” before building.
- •Product recipe: maximize customer value and craft a delightful UX
- •FinTech constraint: you rarely invent new products; you differentiate within rules
- •Exercise: define 3–4 ways your version is “infinitely better”
- •“Swipeys” as a working-backwards artifact to force crisp value props
- 25:10 – 30:46
Turning compliance into a partner: avoiding “Frankenstein” UX in regulated products
They unpack why many FinTechs end up with cluttered, degraded experiences after multiple reviews. Abhishek credits deep domain expertise in legal/compliance, strong culture, and PM practices that build shared vision and reduce adversarial dynamics.
- •Expert legal/compliance teams can enable great UX (not just block)
- •Assume good intent; legal’s job is safe, compliant shipping
- •Get cross-functional buy-in on the product vision early
- •Understand the rule and the true risk; most issues live in gray areas
- 30:46 – 40:12
Scaling Robinhood (2016→): talent density, mission scope, and shipping cadence
Abhishek reflects on why he joined in 2016 and what kept him: exceptional talent density, obsessive product focus, and a massive mission that expands into banking, credit cards, and global ambitions. He also touches on how Robinhood thinks about growth without speculating on stock price.
- •Reasons to join/stay: people, product craft, ambitious mission
- •Broad roadmap: brokerage + money products + global expansion
- •Growth drivers mentioned: retirement match, assets growth, product velocity
- •Market-cap talk is secondary to building toward the next launch/keynote
- 40:12 – 50:40
Org structure after 2022 layoffs: GM model, planning via keynotes, and goal-setting
The conversation gets tactical about how the product org works today. Abhishek explains the shift to business-unit GMs after the 2022 reduction in force, the planning process anchored by big bets and (increasingly) upcoming keynotes, and simplified goal systems instead of formal OKRs.
- •Business units: brokerage, crypto, money (credit card/banking), plus sub-businesses
- •2022 shift to GM structure improved alignment across functions
- •Planning: leadership sets big bets/targets; teams return execution plans
- •Simplified goals over OKRs; quarterly refresh and grading
- 50:40 – 54:40
AI in the PM workflow + PRDs vs prototypes: show the product, not the document
Abhishek shares how he encourages PMs to use AI primarily for early-stage ideation and domain learning, plus reducing repetitive toil. On product reviews, he prefers swipeys/mocks/prototypes over PRDs—especially because customers experience the UI, not the write-up—while acknowledging PRDs remain necessary in FinTech for rules and edge cases.
- •Best AI use cases: ideation, research, rapid learning, admin efficiency
- •AI isn’t yet a replacement for deep product insight or full PRDs
- •In FinTech, PRDs capture rules/regs and edge cases—hard to eliminate
- •Product reviews should center on prototypes/swipeys and real usage
- 54:40 – 1:01:09
Training the team to ship polished products: pixels, dogfooding, and APM program
Abhishek describes what new PMs find most different at Robinhood: a very high bar for pixel-level polish and accountability for the final built experience. He explains how trust with design is built through collaboration, heavy dogfooding, and a willingness to delay launches for quality; he also briefly covers the new APM program setup.
- •High expectation for PM attention to pixel-level details and polish
- •Collaboration principle: ask/design critique vs. stepping on design ownership
- •Dogfooding: internal usage raises the bar; delays are acceptable for quality
- •APM program: early days; experimenting with small rapid wins vs. long projects
- 1:01:09 – 1:05:38
Abhishek’s origin story: referral program experimentation and the “claim your stock” insight
Abhishek recounts joining via an engineering internship referral and moving into product as the company was small and roles blurred. He then gives a rare deep dive into the referral program’s iterative testing—progressing from cash to variable cash to variable stock—and the crucial activation change that made users claim their reward.
- •Entry path: networked internship → full-time; engineering/product blur
- •Early product work: built growth and referral program
- •Experimentation sequence: cash → variable cash → variable stock
- •Key unlock: requiring users to affirmatively claim stock boosted activation
- •Follow-on nudges: notifications and ownership mindset to drive engagement
- 1:05:38 – 1:10:48
Why many FinTech PMs struggle: shallow copying vs. deep hypotheses + cross-functional mastery
Aakash and Abhishek diagnose why many teams don’t reach Robinhood’s iteration velocity: they copy surface-level tactics instead of extracting underlying principles. Abhishek emphasizes treating all regulated-industry partners as core teammates and learning the true constraints to avoid talking past each other.
- •Don’t copy tactics blindly; extract the underlying principle (e.g., “a ride,” not $10)
- •Assume other companies also had imperfect iterations behind the scenes
- •FinTech requires more partners than typical tech: ops, compliance, risk, etc.
- •PM success = rapid learning + aligning intent across functions
- 1:10:48 – 1:18:57
Building “real” money products: design for emotions, plus career lessons and choosing the right company
Abhishek explains how Robinhood designs financial products around meaningful user moments (first stock, saving for retirement) and the emotions they should evoke. He closes with career principles—doing the detailed work, building trust broadly—and how to choose high-upside companies: clear product-market fit, responsible founders, and a culture that invests in people.
- •Define what your product will do better than alternatives, within constraints
- •Design for emotion: make saving/investing feel positive and memorable
- •Career habit: don’t shy away from detail work; show up for the team
- •Choosing companies: strong PMF signals, lean execution, founders who bet on talent
