Snacks Weekly on Data Science

In this episode, we explore how Flipkart rethought ranking in its "Similar Products" widget. Instead of relying on fixed ad slots, the team built a unified ranking system that enables sponsored and organic products to compete on the same playing field. By combining a unified ranking model with a score normalization layer, the system ranks content based on contextual merit rather than pre...

In this episode, we explore how Pinterest rethought its ads serving infrastructure to support more expressive ranking models beyond the traditional two-tower architecture. Their solution was not simply to build a better model, but to redesign the serving stack so that ranking logic and model inference work more closely together, allowing lightweight ranking models to capture richer signals while o...

In this episode, we explore how Meta addressed a fundamental limitation of applying LLMs to enterprise analytics. While modern models are highly capable of generating code and SQL, they still fall short when it comes to organizational context and deep domain understanding — both of which are essential for reliable, real-world analytical work. Meta’s approach focuses on closing this gap through sha...

In this episode, we explore how CVS Health tackled a classic large-scale engineering operations problem: keeping application runbooks accurate, complete, and continuously compliant across hundreds of internal systems. To solve this, the team built an LLM-based automated auditing pipeline. The result is a lightweight but effective system that turns documentation compliance from a periodic manual ef...

In this episode, we explore how Uber tackled one of the most operationally challenging parts of its marketplace: airport pickups. Unlike normal city rides, airport demand is highly bursty, queue-driven, and heavily influenced by flight schedules, delays, and driver positioning decisions. To solve this, Uber built a coordinated forecasting system composed of three specialized models: Estimated Time...

In this episode, we discuss Expedia’s need to evaluate many ranking ideas quickly so that A/B testing would not become a bottleneck. We explore their three-stage experimentation funnel: backtesting to remove weak candidates, interleaving to rapidly compare promising ones, and A/B testing to validate final business impact. What made this design approach effective was using the right evaluation tool...

In this episode, we look at how companies deal with large volumes of unstructured text and why traditional clustering methods often fall short at scale. We explore two LLM-powered approaches shared by data scientists from Microsoft: a bottom-up pipeline that builds structure from data using embeddings and clustering, and a top-down pipeline that starts with LLM-generated categories and refines the...

In this episode, we discuss a classic scaling problem in fraud and risk operations: too much manual review, inconsistent judgments, and growing complexity. We explore the team’s solution, Bumblebee, a multi-agent AI architecture that separates planning, evidence gathering, and analysis into specialized roles, enabling a robust and scalable system to solve the problem. For more details, you can ref...

In this episode, we explore how OLX improved discovery by combining keyword search and vector search instead of forcing a choice between the two. Keyword systems remain excellent for precision, while vector systems add semantic understanding. Together, they create a smarter and more user-friendly marketplace experience. For more details, you can refer to their published tech blog, linked here for...

In this episode, we explore how Udemy built a multilingual AI platform to bring its generative AI features to learners around the world. The team approached localization across three levels: a translation-first approach for broad and fast coverage, a fully native multilingual system for markets where fluency and cultural precision are essential, and a hybrid solution in between that intelligently...

In this episode, we explore how Meta uses the “Ladder of Evidence” framework to evaluate the effectiveness of new product features. Instead of relying on a single analytical method, this framework helps teams choose the right type of evidence based on real-world constraints, leading to better and more informed product decisions. For more details, you can refer to their published tech blog, linked...

In this episode, we explore how Vimeo built a customized AI system for subtitle translation—one that goes beyond basic text translation to tackle the much more challenging problem of synchronizing language with timing. We discuss how the team designed a split-brain architecture to separate translation quality from timing constraints, and how they implemented fallback mechanisms to ensure the syste...

In this episode, we explore how the New York Times engineering team used AI agents to scale unit test coverage across their News site. They accomplished this by building a custom coverage measurement tool, designing a two-loop human–AI workflow, and investing heavily in prompt engineering, including strict guardrails to prevent the agent from cheating or drifting. The key takeaway is that AI works...

In this episode, we explore how Shopify evolved its product classification system across three major stages: from a traditional logistic regression model with TF-IDF features, to a multi-modal approach combining text and images, and finally to Vision Language Models built on top of a standardized and evolving product taxonomy. We also look at how architectural design and inference optimization are...

In this episode, we explore how Faire built its ads system from scratch. On the business side, we discuss why ads matter for a growing marketplace: enabling brand discovery, creating a new revenue stream, and strengthening the overall ecosystem. On the technical side, we break down the three core components—Ads Delivery, Ads Manager, and Ads Foundation—and examine key considerations such as optimi...

In this episode, we explore how Agoda tackled a costly engineering bottleneck by integrating GPT into their CI/CD pipeline to analyze failing SQL stored procedures automatically and suggest optimizations — complete with rewritten queries, index recommendations, and side-by-side performance comparisons. The result is a human-in-the-loop system where AI handles the heavy lifting and engineers make t...

In this episode, we explore how LinkedIn is reimagining job search with AI and large language models — evolving from rigid, keyword-based systems to flexible, intent-aware experiences that feel more conversational and personalized. For more details, you can refer to their published tech blog, linked here for your reference: https://www.linkedin.com/blog/engineering/ai/building-the-next-generation-...

In this episode, we explore how Uber tackled the challenge of personalizing CRM communications at scale through contextual bandit strategies enhanced with generative AI embeddings, lightweight and powerful models like LinUCB and XGBoost, and smart decision augmentation with SquareCB. This work shows how data science can take a core business need—delivering relevant user communications—and build sy...

In this episode, we explore how seemingly perfect-looking SQL generated by AI agents can be “lying” when essential logic is missing. The Thomson Reuters Labs team highlights the need for deeper evaluation beyond simple syntax checks, and shows how tools like TruLens and AgentBench help expose hidden errors and better align agent outputs with real business intent. For more details, you can refer to...

In this episode, we explore how Airbnb measures Listing Lifetime Value by separating it into baseline LTV, incremental LTV, and marketing-induced incremental LTV, and how this framework helps address challenges like measuring true incrementality and handling uncertainty about the future. For more details, you can refer to their published tech blog, linked here for your reference: https://medium.co...

In this episode, we explore how OLX evolved its ranking algorithms—from the pairwise logic of RankNet to the metric-optimized power of LambdaRank—to ensure users find exactly what they’re looking for. We discuss how moving from simple classification to "Learning to Rank" helps businesses prioritize user attention where it matters most. For more details, you can refer to their published t...

In this episode, we explore how Udemy tackled the tricky challenge of understanding learner intent in their AI Assistant — from a simple similarity-based embedding model, through experiments with larger models and fine-tuning, to a hybrid system that intelligently leverages both embeddings and large language model classification. This evolution demonstrates how real-world ML systems often require...

In this episode, we explore how Meta’s data scientists approach product strategy using a structured framework that adapts to different data and problem scenarios. We walk through the distinct analytical approaches used across different problem spaces, defined by whether data availability is high or low and whether problem clarity is broad or concrete. Each scenario requires a different mix of thin...

In this episode, we explore how PayPal estimates incremental lift in customer value using synthetic control methods. This causal inference–based approach provides a principled way to construct a counterfactual and isolate causal effects when traditional experiments aren’t sufficient, helping teams measure true impact in a complex, noisy, real-world environment and make more informed decisions. For...

In this episode, we explore how Netflix tackles the challenge of predicting user session intent by extending the capabilities of its foundation model with a hierarchical multi-task learning architecture. This approach helps Netflix better understand what users want in the moment and personalize the experience in real time, ultimately improving its recommendation system at scale. For more details,...