No Priors Ep. 50 | With Stripe Head of Information Emily Glassberg Sands

Many companies that are building AI products for their users are not primarily AI companies. Today on No Priors, Sarah and Elad are joined by Emily Glassberg Sands who is the Head of Information at Stripe. They talk about how Stripe prioritizes AI projects and builds these tools from the inside out. Stripe was an early adopter of utilizing LLMs to help their end user. Emily talks about how they decided it was time to meaningfully invest in AI given the trajectory of the industry and the wealth of information Stripe has access to. The company’s goal with utilizing AI is to empower non-technical users to code using natural language and for technical users to be able to work much quicker and in this episode she talks about how their Radar Assistant and Sigma Assistant achieve those goals. Sign up for new podcasts every week. Email feedback to show@no-priors.com Follow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @emilygsands Show Notes: 0:00 Background 0:38 Emily’s role at Stripe 2:31 Adopting early gen AI models 4:44 Promoting internal usage of AI 8:17 Applied ML accelerator teams 10:36 Radar fraud assistant 13:30 Sigma assistant 14:32 How will AI affect Stripe in 3 years 17:00 Knowing when it’s time to invest more fully in AI 18:28 Deciding how to proliferate models 22:04 Whitespace for fintechs employing AI 25:41 Leveraging payments data for customers 27:51 Labor economics and data 30:10 Macro economic trends for strategic decisions 32:54 How will AI impact education 35:36 Unique needs of AI startups

Elad GilhostEmily Glassberg SandsguestSarah Guohost

Feb 7, 202439mWatch on YouTube ↗

WHAT IT’S REALLY ABOUT

Stripe’s Data Chief on Generative AI, Fintech Infrastructure, and Education

Emily Glassberg Sands, Head of Information at Stripe, explains how Stripe is using data science and generative AI to power both internal operations and user-facing financial products. She describes Stripe’s bottoms‑up approach to LLM adoption, starting with an internal LLM Explorer and accelerator teams that seed new AI bets. Sands details concrete applications like Radar Assistant and Sigma Assistant that translate natural language into code and analytics, democratizing fraud controls and business insights for non-technical users. She also connects Stripe’s macroeconomic vantage point and her labor economics background to broader opportunities in fintech, AI-native startups, and the future of education and skills.

IDEAS WORTH REMEMBERING

5 ideas

Start AI adoption with safe, internal experimentation at scale.

Stripe launched an internal LLM Explorer, quickly adopted by ~half the company weekly, to let employees across functions discover practical use cases before heavily committing to external AI features.

Use small, time-bounded accelerator teams to seed new AI bets.

Stripe funds ‘one- to two‑pizza’ accelerator teams for ~six months with clear but flexible charters, enabling focused experimentation (like LLM Explorer and support tools) without disrupting core product roadmaps.

Democratize complex capabilities by turning natural language into code and rules.

Radar Assistant and Sigma Assistant let non-technical users express fraud policies and analytics questions in plain English, which Stripe converts into executable rules and SQL, enlarging who can act on data and risk.

Centralize AI infrastructure but decentralize model choice.

Stripe provides a shared LLM API, security, and experimentation platform, sets sensible default models, and then lets product teams trade off cost, latency, and quality—internally billing high-usage apps to enforce discipline.

Leverage proprietary payments data for optimization and future ‘economic OS’ products.

Existing ML already improves conversion, auth rates, fraud detection, and dunning; Sands envisions future foundation models on financial data that could meaningfully boost performance and automate pricing, discounting, and geo strategy.

WORDS WORTH SAVING

5 quotes

We were looking at the technical breakthroughs and the product launches all over the ecosystem with awe, but also honestly, a little bit of overwhelm.

— Emily Glassberg Sands

The weekly active user count of this LLM Explorer is still at almost 3,000, which is just shy of half the company using it every single week.

— Emily Glassberg Sands

We’re very focused on leveraging AI so that non-technical folks that are users can do things that they couldn’t do before, and so that technical folks can move an order of magnitude faster.

— Emily Glassberg Sands

It doesn’t feel crazy to think that a good foundation model could outperform more traditional approaches by, I don’t know, 100 bips, 200 bips.

— Emily Glassberg Sands

Whether it’s Stripe or someone else, using financial data to help businesses be more successful, to grow the pie, to grow the GDP, I think is really powerful.

— Emily Glassberg Sands

Stripe’s Information organization structure and responsibilitiesInternal LLM adoption via tools like LLM Explorer and accelerator teamsUser-facing AI products: Radar Assistant and Sigma AssistantAI infrastructure choices: model selection, experimentation, and central platformsFintech and AI opportunities: payments optimization, economic operating system visionDistinct needs and patterns of AI-native startups on StripeAI and education: skills, credentials, and labor market signaling

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.