Techlist.io - Korean Tech Blog Curator

googleOriginal article

A differentially private framework for gaining insights into AI chatbot use (opens in new tab)

Google Research has introduced Urania, a novel framework designed to extract high-level usage insights from AI chatbot conversations while maintaining rigorous differential privacy (DP) guarantees. Unlike previous heuristic methods that rely on simple redaction or LLM-based PII stripping, this pipeline ensures that no individual user's data can be reconstructed from the resulting summaries. By combining DP clustering and keyword extraction with LLM-based summarization, the system provides a formal, auditable approach to understanding platform trends without compromising sensitive information. ## Limitations of Heuristic Privacy * Existing frameworks often rely on large language models to manually strip personally identifiable information (PII) from text before analysis. * These heuristic protections are difficult to formalize or audit, and their effectiveness may diminish as models evolve or face sophisticated prompt injection attacks. * The Urania framework addresses these weaknesses by using mathematical privacy budgets (the epsilon parameter) to measure and limit the influence of any single user's data on the final output. ## The Differentially Private Pipeline * **DP Clustering**: The framework first converts conversation data into numerical embeddings. These are grouped using a DP clustering algorithm, ensuring that cluster centers reflect broad trends rather than specific individual inputs. * **DP Keyword Extraction**: The system identifies keywords for each cluster and generates a histogram of their frequency. By adding mathematical noise to these counts, the framework masks individual contributions and ensures that only keywords common to many users are retained. * **Keyword Generation Methods**: The researchers explored three methods for extraction: LLM-guided selection of relevant terms, a differentially private version of TF-IDF, and an LLM-guided approach that selects terms from a pre-defined list of public keywords. * **LLM Summarization**: In the final stage, an LLM generates a high-level summary of the cluster using only the noisy, anonymized keywords. Because the LLM never sees the raw conversation text, the "post-processing" property of DP guarantees that the final summary remains private. ## Privacy and Utility Trade-offs * The framework was tested against a non-private baseline (Simple-CLIO) to evaluate how privacy constraints affect the quality of the insights generated. * Stronger privacy settings (lower epsilon values) inherently result in a utility trade-off, as the added noise can obscure some niche usage patterns. * Despite these trade-offs, the framework provides a robust defense against data leakage, as the summarization model is structurally prevented from seeing sensitive original text, making it resilient to prompt injection. This framework offers a scalable way for platform providers to analyze chatbot usage patterns and enforce safety policies while providing mathematical certainty regarding user privacy. For organizations handling sensitive conversation data, moving from heuristic redaction to formal DP pipelines like Urania provides a more robust and auditable path for service improvement.

figma2 min readCurated summary

Introducing Three New Tools For Precise Image Editing In Figma | Figma Blog

Figma introduced three AI-powered image editing tools—Erase object, Isolate object, and Expand image—to make detailed image manipulation possible without leaving the design canvas. The tools complement existing features such as background removal, cropping, and AI image generation, helping designers refine assets in context. A new image-editing toolbar brings these capabilities together for faster, more integrated workflows. ## Erase and Isolate Objects - **Erase object** removes a selected object from an image. - **Isolate object** separates an object or person so it can be edited or repositioned without changing the background. - Users can select objects with a lasso and apply: - Lighting and color adjustments - Blur and focus effects - Color correction - Shadows - These tools are useful for refining product photos, removing distractions, and emphasizing key visual elements. - Text-prompt editing remains available through Figma’s **Edit image** feature, but the new tools provide more precise manual control. ## Expand Images for New Layouts - **Expand image** generates additional background content to fit a new aspect ratio. - It adapts images for formats such as: - Mobile layouts - Web banners - Social media assets - Unlike cropping, expansion preserves the original subject and surrounding context without distortion. - For example, a square product image can be expanded into a wide banner while leaving room for text. ## A Unified Image-Editing Toolbar - The new toolbar combines the three AI tools with existing capabilities, including: - Remove background - Crop - AI image generation and editing - Remove background is now easier to find because it is one of the most frequently used AI actions in Figma. - The tools are available across Figma, including FigJam, Slides, and Buzz beta, with some seat restrictions. - In Figma Design and Figma Draw, they are available to Full-seat users on Professional, Organization, and Enterprise plans with AI enabled. - AI actions consume Figma credits. These updates aim to keep image editing inside Figma, reducing the need to switch between external tools. Designers can now make precise object-level edits, adapt images to different formats, and maintain visual consistency directly within their workflows.

Read original(opens in new tab)
awsOriginal article

AWS Weekly Roundup: AWS re:Invent keynote recap, on-demand videos, and more (December 8, 2025) (opens in new tab)

The December 8, 2025, AWS Weekly Roundup recaps the major themes from AWS re:Invent, signaling a significant industry transition from AI assistants to autonomous AI agents. While technical innovation in infrastructure remains a priority, the event underscored that developers remain at the heart of the AWS mission, empowered by new tools to automate complex tasks using natural language. This shift represents a "renaissance" in cloud computing, where purpose-built infrastructure is now designed to support the non-deterministic nature of agentic workloads. ## Community Recognition and the Now Go Build Award * Raphael Francis Quisumbing (Rafi) from the Philippines was honored with the Now Go Build Award, presented by Werner Vogels. * A veteran of the ecosystem, Quisumbing has served as an AWS Hero since 2015 and has co-led the AWS User Group Philippines for over a decade. * The recognition emphasizes AWS's continued focus on community dedication and the role of individual builders in empowering regional developer ecosystems. ## The Evolution from AI Assistants to Agents * AWS CEO Matt Garman identified AI agents as the next major inflection point for the industry, moving beyond simple chat interfaces to systems that perform tasks and automate workflows. * Dr. Swami Sivasubramanian highlighted a paradigm shift where natural language serves as the primary interface for describing complex goals. * These agents are designed to autonomously generate plans, write necessary code, and call various tools to execute complete solutions without constant human intervention. * AWS is prioritizing the development of production-ready infrastructure that is secure and scalable specifically to handle the "non-deterministic" behavior of these AI agents. ## Core Infrastructure and the Developer Renaissance * Despite the focus on AI, AWS reaffirmed that its core mission remains the "freedom to invent," keeping developers central to its 20-year strategy. * Leaders Peter DeSantis and Dave Brown reinforced that foundational attributes—security, availability, and performance—remain the non-negotiable pillars of the AWS cloud. * The integration of AI agents is framed as a way to finally realize material business returns on AI investments by moving from experimental use cases to automated business logic. To maximize the value of these updates, organizations should begin evaluating how to transition from simple LLM implementations to agentic frameworks that can execute end-to-end business processes. Reviewing the on-demand keynote sessions from re:Invent 2025 is recommended for technical teams looking to implement the latest secure, agent-ready infrastructure.

tossOriginal article

Improving Business Data Literacy: (opens in new tab)

Toss’s Business Data Team addressed the lack of centralized insights into their business customer (BC) base by building a standardized Single Source of Truth (SSOT) data mart and an iterative Monthly BC Report. This initiative successfully unified fragmented data across business units like Shopping, Ads, and Pay, enabling consistent data-driven decision-making and significantly raising the organization's overall data literacy. ## Establishing a Single Source of Truth (SSOT) - Addressed the inefficiency of fragmented data across various departments by integrating disparate datasets into a unified, enterprise-wide data mart. - Standardized the definition of an "active" Business Customer through cross-functional communication and a deep understanding of how revenue and costs are generated in each service domain. - Eliminated communication overhead by ensuring all stakeholders used a single, verified dataset rather than conflicting numbers from different business silos. ## Designing the Monthly BC Report for Actionable Insights - Visualized monthly revenue trends by segmenting customers into specific tiers and categories, such as New, Churn, and Retained, to identify where growth or attrition was occurring. - Implemented Cohort Retention metrics by business unit to measure platform stickiness and help teams understand which services were most effective at retaining business users. - Provided granular Raw Data lists for high-revenue customers showing significant growth or churn, allowing operational teams to identify immediate action points. - Refined reporting metrics through in-depth interviews with Product Owners (POs), Sales Leaders, and Domain Heads to ensure the data addressed real-world business questions. ## Technical Architecture and Validation - Built the core SSOT data mart using Airflow for scalable data orchestration and workflow management. - Leveraged Jenkins to handle the batch processing and deployment of the specific data layers required for the reporting environment. - Integrated Tableau with SQL-based fact aggregations to automate the monthly refresh of charts and dashboards, ensuring the report remains a "living" document. - Conducted "collective intelligence" verification meetings to check metric definitions, units, and visual clarity, ensuring the final report was intuitive for all users. ## Driving Organizational Change and Data Literacy - Sparked a surge in data demand, leading to follow-up projects such as daily real-time tracking, Cross-Domain Activation analysis, and deeper funnel analysis for BC registrations. - Transitioned the organizational culture from passive data consumption to active utilization, with diverse roles—including Strategy Managers and Business Marketers—now using BC data to prove their business impact. - Maintained an iterative approach where the report format evolves every month based on stakeholder feedback, ensuring the data remains relevant to the shifting needs of the business. Establishing a centralized data culture requires more than just technical infrastructure; it requires a commitment to iterative feedback and clear communication. By moving from fragmented silos to a unified reporting standard, data analysts can transform from simple "number providers" into strategic partners who drive company-wide literacy and growth.

naverOriginal article

When Design Systems Meet AI: Shifts (opens in new tab)

The integration of AI into the frontend development workflow is transforming how markup is generated, shifting the developer's role from manual coding to system orchestration. By leveraging Naver Financial’s robust design system—comprised of standardized design tokens and components—developers can use AI to automate the translation of Figma designs into functional code. This evolution suggests a future where the efficiency of UI implementation is dictated by the maturity of the underlying design system and the precision of AI instructions. ### Foundations of the Naver Financial Design System * The system is built on "Design Tokens," which serve as the smallest units of design, such as colors, typography, and spacing, ensuring consistency across all platforms. * Pre-defined components act as the primary building blocks for the UI, allowing the AI to reference established patterns rather than generating arbitrary styles. * The philosophy of "knowing your system" is emphasized as a prerequisite; AI effectiveness is directly proportional to how well-structured the design assets and code libraries are. ### Automating Markup with Code Connect and AI * Figma's "Code Connect" is utilized to bridge the gap between design files and the actual codebase, providing a source of truth for how components should be implemented. * Specific "Instructions" or prompts are developed to guide the AI in mapping Figma properties to specific React component props and design system logic. * This approach enables the transition from "drawing" UI to "declaring" it, where the AI interprets the design intent and outputs code that adheres to the organization’s technical standards. ### Challenges and Limitations in Real-World Development * While AI-generated markup provides a strong starting point, it often requires manual intervention for complex business logic, state management, and edge-case handling. * Maintaining the "Instruction" set requires ongoing effort to ensure the AI stays updated with the latest changes in the component library. * Developers must transition into a "reviewer" role, as the AI can still struggle with the specific context of a feature or integration with legacy code structures. The path to fully automated frontend development requires a highly mature design system as its backbone. For teams looking to adopt this paradigm, the priority should be standardizing design tokens and component interfaces; only then can AI effectively reduce the "last mile" of markup work and allow developers to focus on higher-level architectural challenges.

figma2 min readCurated summary

Updates to AI Credits in Figma | Figma Blog

Figma is expanding access to AI while adding clearer controls over credit consumption. The company has introduced usage tracking for both administrators and individual users, and will soon offer subscription-based and pay-as-you-go options for additional credits. Seat credit limits will become enforceable beginning March 18, 2026. ## Tracking AI Credit Usage - Billing administrators can see: - Which team members use AI features - How many credits each person has consumed - Which users have run out of credits - How many days remain before credits reset - Individual users can view their remaining balance and reset date. - Most AI actions will display the number of credits they consume. ## Purchasing Additional Credits Beginning March 11, 2026, teams will have two ways to buy credits beyond their included allowances: - **AI credits subscription** - Provides a shared monthly credit pool. - Offers better rates for teams with regular AI usage. - Packages begin at 5,000 credits per month. - **Pay-as-you-go billing** - Charges teams only when additional credits are needed. - Can be used independently or alongside a subscription. - Supports a team-defined spending limit. - Expected to become available by Q2 2026. Example pricing includes: - 5,000 credits: $120 monthly subscription or $150 pay-as-you-go - 7,500 credits: $180 monthly subscription or $225 pay-as-you-go - 10,000 credits: $240 monthly subscription or $300 pay-as-you-go ## Included Credit Allowances and Enforcement Monthly seat allowances vary by plan: - Starter: 500 credits - Professional full seats: 3,000 credits - Organization full seats: 3,500 credits - Enterprise full seats: 4,250 credits - Other seats: generally 500 credits Figma will begin enforcing all seat credit limits on March 18, 2026. Teams should therefore monitor usage and choose either an add-on subscription or pay-as-you-go billing if their AI needs exceed the included amounts.

Read original(opens in new tab)
pinterest4 min readCurated summary

How Pinterest Built a Real‑Time Radar for Violative Content using AI

Pinterest built an AI-assisted prevalence measurement system to estimate how often users actually see policy-violating content, rather than relying only on user reports. The system samples daily impressions, uses production risk scores to improve efficiency, labels content with a multimodal LLM, and applies statistical reweighting to preserve unbiased estimates. This enables daily, segmented monitoring with substantially lower cost and latency than human-only review. ## Why Prevalence Matters - User reports miss important harms because: - Some sensitive issues, such as self-harm, are under-reported. - Users seeking harmful content may not report it. - Rare policy categories provide too few reports for reliable trend detection. - Human review of reports is expensive and slow. - Prevalence measures exposure: the share of total views directed to violating content. - This helps Pinterest identify under-reported harms, evaluate interventions, and detect changes earlier. - Human-only prevalence studies were previously conducted only about every six months and required multiple reviewers plus adjudication. ## What Pinterest Measures - Daily prevalence is calculated as: - **Views of content violating a policy ÷ total views** - For example, 10 violating views in a sample of 100,000 produces an estimated prevalence of 0.01%. - Results include 95% confidence intervals to communicate statistical precision. - Metrics can be segmented by: - Policy area, such as Adult Content, Self-harm, or Graphic Violence - Sub-policy, such as nudity versus explicit sexual content - Surface, including Homefeed, Search, and Related Pins - Content age, geography, and user-age groups where relevant ## Risk-Aware, Unbiased Sampling - Pinterest samples from the daily user-impressions stream. - Production enforcement risk scores are used to prioritize likely high-risk and high-exposure content, but they are not treated as labels or eligibility rules. - Missing scores are replaced with the day’s median so that new content remains eligible. - Weighted reservoir sampling approximates probability-proportional-to-size sampling, considering impressions and risk scores. - Inverse-probability weighting removes the bias introduced by risk-based sampling, ensuring estimates represent impressions rather than model thresholds. - Pinterest uses Hansen–Hurwitz ratio estimators for sampling with replacement and Horvitz–Thompson ratio estimators for sampling without replacement. - Pure random sampling is also available for validation studies. ## LLM-Based Labeling - A multimodal LLM analyzes sampled content using both images and text. - Prompts are reviewed by policy subject-matter experts and can return structured label hierarchies such as `safe`, `not_safe`, and `unsure`. - Each decision records: - The label and brief rationale - Policy version - Prompt and model identifiers - Token usage and run cost - Human validation is performed on strategically selected samples to identify edge cases and AI blind spots. - The LLM is tested against human-reviewed gold sets before launch and periodically afterward to detect drift. - The workflow is reportedly 15 times faster and far cheaper than human-only labeling while maintaining comparable decision quality and statistical governance. ## Production System and Monitoring - Inputs include entity-by-day engagement data such as impressions, clicks, hides, and reports, alongside current production risk scores. - The system stores prevalence estimates, sampling weights, labels, diagnostics, and lineage for audits. - Dashboards display: - Daily prevalence and 95% confidence intervals - Confidence-interval width and effective sample size - Sample positive rate - Risk-score distributions - Prompt, model, taxonomy, and metric versions - Teams can pivot results by policy, sub-policy, and surface. - Validation samples and run-health information help monitor both statistical quality and operational reliability. Pinterest’s approach combines probability sampling, inverse-probability estimation, and continuously calibrated multimodal AI labeling to create a daily radar for harmful exposure. The practical recommendation is to use AI to scale measurement, but retain rigorous sampling, human validation, confidence intervals, and full model and policy lineage so that faster estimates remain trustworthy.

Read original(opens in new tab)
daangnOriginal article

No Need to Fetch Everything Every Time (opens in new tab)

To optimize data synchronization and ensure production stability, Daangn’s data engineering team transitioned their MongoDB data pipeline from a resource-intensive full-dump method to a Change Data Capture (CDC) architecture. By leveraging Flink CDC, the team successfully reduced database CPU usage to under 60% while consistently meeting a two-hour data delivery Service Level Objective (SLO). This shift enables efficient, schema-agnostic data replication to BigQuery, facilitating high-scale analysis without compromising the performance of live services. ### Limitations of Traditional Dump Methods * The previous Spark Connector-based approach required full table scans, leading to a direct trade-off between hitting delivery deadlines and maintaining database health. * Increasing data volumes caused significant CPU spikes, threatening the stability of transaction processing in production environments. * Standard incremental loads were unreliable because many collections lacked consistent `updated_at` fields or required the tracking of hard deletes, which full dumps handle poorly at scale. ### Advantages of Flink CDC for MongoDB * Flink CDC provides native support for MongoDB Change Streams, allowing the system to read the Oplog directly and use resume tokens to restart from specific failure points. * The framework’s checkpointing mechanism ensures "Exactly-Once" processing by periodically saving the pipeline state to distributed storage like GCS or S3. * Unlike standalone tools like Debezium, Flink allows for an integrated "Extract-Transform-Load" (ETL) flow within a single job, reducing operational complexity and the need for intermediate message queues. * The architecture is horizontally scalable, meaning TaskManagers can be increased to handle sudden bursts in event volume without re-architecting the pipeline. ### Pipeline Architecture and Processing Logic * The core engine monitors MongoDB write operations (Insert, Update, Delete) in real-time via Change Streams and transmits them to BigQuery. * An hourly batch process is utilized rather than pure real-time streaming to prioritize operational stability, idempotency, and easier recovery from failures. * The downstream pipeline includes a Schema Evolution step that automatically detects and adds new fields to BigQuery tables, ensuring the NoSQL-to-SQL transition is seamless. * Data processing involves deduplicating recent change events and merging them into a raw JSON table before materializing them into a final structured table for end-users. For organizations managing large-scale MongoDB clusters, implementing Flink CDC serves as a powerful solution to balance analytical requirements with database performance. Prioritizing a robust, batch-integrated CDC flow allows teams to meet strict delivery targets and maintain data integrity without the infrastructure overhead of a fully real-time streaming system.

daangnOriginal article

Why fetch it all every (opens in new tab)

As Daangn’s data volume grew, their traditional full-dump approach using Spark for MongoDB began causing significant CPU spikes and failing to meet the two-hour data delivery Service Level Objectives (SLOs). To resolve this, the team implemented a Change Data Capture (CDC) pipeline using Flink CDC to synchronize data efficiently without the need for resource-intensive full table scans. This transition successfully stabilized database performance and ensured timely data availability in BigQuery by focusing on incremental change logs rather than repeated bulk extracts. ### Limitations of Traditional Dump Methods * The previous Spark Connector method required full table scans, creating a direct conflict between service stability and data freshness. * Attempts to lower DB load resulted in missing the 2-hour SLO, while meeting the SLO pushed CPU usage to dangerous levels. * Standard incremental loading was ruled out because it relied on `updated_at` fields, which were not consistently updated across all business logic or schemas. * The team targeted the top five largest and most frequently updated collections for the initial CDC transition to maximize performance gains. ### Advantages of Flink CDC * Flink CDC provides native support for MongoDB Change Streams, allowing the system to use resume tokens and Flink checkpoints for seamless recovery after failures. * It guarantees "Exactly-Once" processing by periodically saving the pipeline state to distributed storage, ensuring data integrity during restarts. * Unlike tools like Debezium that require separate systems for data processing, Flink handles the entire "Extract-Transform-Load" (ETL) lifecycle within a single job. * The architecture is horizontally scalable; increasing the number of TaskManagers allows the pipeline to handle surges in event volume with linear performance improvements. ### Pipeline Architecture and Implementation * The system utilizes the MongoDB Oplog to capture real-time write operations (inserts, updates, and deletes) which are then processed by Flink. * The backend pipeline operates on an hourly batch cycle to extract the latest change events, deduplicate them, and merge them into raw JSON tables in BigQuery. * A "Schema Evolution" step automatically detects and adds missing fields to BigQuery tables, bridging the gap between NoSQL flexibility and SQL structure. * While Flink captures data in real-time, the team opted for hourly materialization to maintain idempotency, simplify error recovery, and meet existing business requirements without unnecessary architectural complexity. For organizations managing large-scale MongoDB instances, moving from bulk extracts to a CDC-based model is a critical step in balancing database health with analytical needs. Implementing a unified framework like Flink CDC not only reduces the load on operational databases but also simplifies the management of complex data transformations and schema changes.

woowahanOriginal article

In Search of Lost Accessibility | Woowa (opens in new tab)

Achieving a high accessibility score on automated tools like Lighthouse does not always translate to a functional experience for users with visual impairments. This post explores how a team discovered that their "high-scoring" product actually required over 300 swipes for a screen reader user to reach a purchase button, leading them to overhaul their approach. By focusing on actual screen reader behavior rather than just checklists, they successfully transformed a fragmented interface into a streamlined, navigable user journey. ### Navigational Structure with Landmarks and Headings * The team implemented a clear hierarchy using landmarks (header, main, footer) and heading levels, which allows screen reader users to jump between sections via tools like the iOS VoiceOver "Rotor." * To ensure consistency, they developed a reusable component that automatically wraps content in a `<section>` and links it to a heading using the `aria-labelledby` attribute. * They addressed a common CSS pitfall: because setting `list-style: none` can cause VoiceOver to stop recognizing elements as a list, they explicitly added `role="list"` to maintain structural context for the user. ### Consolidating Fragmented Text for Readability * Information that should be heard as a single unit, such as prices (e.g., "990" and "Won"), was often fragmented into separate swipes; the team corrected this by using template literals to merge data into single strings. * For cases where visual styling required separate DOM elements, they used a "NoScreen" component strategy: hiding the visual elements from screen readers with `aria-hidden="true"` while providing a single, visually hidden description for the screen reader to announce. * The team noted that `aria-label` on generic containers like `<span>` or `<div>` is often ignored by iOS VoiceOver, making screen-reader-only text a more reliable method for cross-platform accessibility. ### Defining Roles for Interactive Elements * The team identified that generic buttons like "View All" lacked context, so they updated them with specific labels (e.g., "View all 20 reviews") to clarify the outcome of the interaction. * They ensured that all interactive elements have clearly defined roles, preventing the ambiguity that occurs when a screen reader identifies an element as a "button" without explaining its specific purpose or the data it controls. True accessibility is best measured by the physical effort required to complete a task, such as the number of swipes or touches. Developers should move beyond automated audits and regularly perform manual testing with screen readers like VoiceOver or TalkBack to ensure their services are genuinely usable for everyone.

woowahanOriginal article

아한형제들 기술블로그 (opens in new tab)

The 7th Woowacourse crew has successfully launched three distinct services, demonstrating that modern software engineering requires a synergy of technical mastery and "soft skills" like product planning and team communication. By owning the entire lifecycle from ideation to deployment, these developers moved beyond mere coding to solve real-world problems through agile iterations, user feedback, and robust infrastructure management. The program’s focus on the full stack of development—including monitoring, 2-week sprints, and collaborative design—highlights a shift toward producing well-rounded engineers capable of navigating professional environments. ### The Woowacourse Full-Cycle Philosophy * The 10-month curriculum emphasizes soft skills, including speaking and writing, alongside traditional technical tracks like Web Backend, Frontend, and Mobile Android. * During Level 3 and 4, crews transition from fundamental programming to managing team projects where they must handle everything from initial architecture to UI/UX design. * The process mimics real-world industry standards by implementing 2-week development sprints, establishing monitoring environments, and managing automated deployment pipelines. * The core goal is to shift the developer's mindset from simply writing code to understanding why certain features are planned and how architecture choices impact the final user value. ### Pickeat: Collaborative Dining Decisions * This service addresses "decision fatigue" during group meals by providing a collaborative platform to filter restaurants based on dietary constraints and preferences. * Technical challenges included frequent domain restructuring and UI overhauls as the team pivoted based on real-world user feedback during demo days. * The platform utilizes location data for automatic restaurant lookups and supports real-time voting mechanisms to ensure democratic and efficient group decisions. * Development focused on aligning team judgment standards and iterating quickly to validate product-market fit rather than adhering strictly to initial specifications. ### Bottari: Real-Time Synchronized Checklists * Bottari is a checklist service designed for situations like traveling or moving, focusing on "becoming a companion for the user’s memory." * The service features template-based list generation and a "Team Bottari" function that allows multiple users to collaborate on a single list with real-time synchronization. * A major technical focus was placed on the user experience flow, specifically optimizing notification timing and sync states to provide "peace of mind" for users. * The project demonstrates the principle that technology serves as a tool for solving psychological pain points, such as the anxiety of forgetting essential items. ### Coffee Shout: Real-Time Betting and Mini-Games * Designed to gamify office culture, this service replaces simple "rock-paper-scissors" with interactive mini-games and weighted roulette for coffee bets. * The technical stack involved challenging implementations of WebSockets and distributed environments to handle the concurrency required for real-time gaming. * The team focused on algorithm balancing for the weighted roulette system to ensure fairness and excitement during the betting process. * Refinement of the service was driven by direct feedback from other Woowacourse crews, emphasizing the importance of community testing in the development lifecycle. These projects underscore that the transition from a student to a professional developer is defined by the ability to manage shifting requirements and technical complexity while maintaining a focus on the end-user's experience.

lineOriginal article

Introducing a New A/B Testing System (opens in new tab)

LY Corporation has developed an advanced A/B testing system that moves beyond simple random assignment to support dynamic user segmentation. By integrating a dedicated targeting system with a high-performance experiment assigner, the platform allows for precise experiments tailored to specific user characteristics and behaviors. This architecture enables data-driven decisions that are more relevant to localized or specialized user groups rather than relying on broad averages. ## Limitations of Traditional A/B Testing * General A/B test systems typically rely on random assignment, such as applying a hash function to a user ID (`hash(id) % 2`), which is simple and cost-effective. * While random assignment reduces selection bias, it is insufficient for hypotheses that only apply to specific cohorts, such as "iOS users living in Osaka." * Advanced systems solve this by shifting from general testing across an entire user base to personalized testing for specific segments. ## Architecture of the Targeting System * The system processes massive datasets including user information, mobile device data, and application activity stored in HDFS. * Apache Spark is used to execute complex conditional operations—such as unions, intersections, and subtractions—to refine user segments. * Segment data is written to Object Storage and then cached in Redis using a `{user_id}-{segment_id}` key format to ensure low-latency lookups during live requests. ## A/B Test Management and Assignment * The system utilizes "Central Dogma" as a configuration repository where operators and administrators define experiment parameters. * A Test Group Assigner orchestrates the process: when a client makes a request, the assigner retrieves experiment info and checks the user's segment membership in Redis. * Once a user is assigned to a specific group (e.g., Test Group 1), the system serves the corresponding content and logs the event to a data store for dashboard visualization and analysis. ## Strategic Use Cases and Future Plans * **Content Recommendation:** Testing different Machine Learning models to see which performs better for a specific user demographic. * **Targeted Incentives:** Limiting shopping discount experiments to "light users," as coupons may not significantly change the behavior of "heavy users." * **Onboarding Optimization:** Restricting UI tests to new users only, ensuring that existing users' experiences remain uninterrupted. * **Platform Expansion:** Future goals include building a unified admin interface for the entire lifecycle of an experiment and expanding the system to cover all services within LY Corporation. For organizations looking to optimize user experience, transitioning from random assignment to dynamic segmentation is essential for high-precision product development. Ensuring that segment data is cached in a high-performance store like Redis is critical to maintaining low latency when serving experimental variations in real-time.

pinterest3 min readCurated summary

Improving Quality of Recommended Content through Pinner Surveys

Pinterest uses Pinner surveys to measure visual quality and incorporate user preferences into recommendation systems, rather than optimizing solely for engagement. The company surveyed 5,000 Pins, trained a lightweight neural network to predict average perceived quality, and applied the resulting model across Homefeed, Related Pins, and Search. This approach aims to reduce clickbait and promote content that supports positive, long-term user experiences. ## Why Engagement Alone Is Insufficient - High engagement does not necessarily indicate high-quality content; optimizing for clicks can promote clickbait or harmful material. - Pinterest defines quality as content that feels good, inspires further exploration, and encourages fulfilling long-term engagement. - Direct user feedback helps recommendation systems prioritize content that Pinners actually value. - The work supports Pinterest’s Inspired Internet Pledge principles, especially listening to users and tuning the platform for wellbeing. ## Collecting Pinner Quality Ratings - Pinners rated images from 1 to 5 in response to: “How visually pleasing or displeasing is this Pin?” - Pinterest collected ratings for 5,000 Pins, sampling 1,000 from each of five major interest categories: - Art - Beauty - DIY & Crafts - Home Decor - Women’s Fashion - Pins were sampled based on impressions and were generally mid-to-high quality rather than deliberately exposing users to poor content. - Each image received at least 10 ratings, allowing Pinterest to average responses and reduce noise from subjectivity or accidental misclicks. - Surveys were considered appropriate for visual appeal, which is subjective but still measurable across many users. More objective issues should be evaluated by trained reviewers, while highly contextual judgments such as personal relevance are harder to capture with a single Pin-level score. - Highly rated content included makeup, grooming styles, maximalist interiors, landscapes, sunsets, and baby animals. - Home Decor images tended to receive higher ratings overall, while Art showed the greatest variation, reflecting its subjective nature. ## Training a Visual-Quality Model - Pinterest trained a model to estimate the average Pinner’s perception of visual quality from image embeddings. - Embeddings encode visual, textual, and behavioral information, including relationships between images and the boards where they are saved. - The model produces a score from 0 to 1, with higher values representing greater perceived quality. - Pinterest chose a small fully connected neural network with approximately 92,000 parameters: - The limited size helps prevent overfitting to the 5,000-image dataset. - It also makes large-scale inference faster and less expensive. - Instead of predicting an exact rating, the model uses pairwise ranking: - It learns which of two images Pinners would consider better. - The comparison is based on each image’s mean survey rating. - Training comparisons are restricted to images within the same top-level interest category, encouraging the model to learn visual quality rather than simply recognizing that one topic is more popular than another. Pinterest’s approach demonstrates how survey-based quality signals can complement engagement metrics. Training recommendation systems on what users perceive as appealing can help the platform promote more satisfying content while reducing incentives to favor attention-grabbing but low-quality material.

Read original(opens in new tab)
netflixOriginal article

AV1 — Now Powering 30% of Netflix Streaming | by Netflix Technology Blog | Dec, 2025 | Netflix TechBlog (opens in new tab)

Netflix has successfully integrated the AV1 codec into its streaming infrastructure, where it now accounts for 30% of all viewing traffic and is on track to become the platform's primary format. This transition from legacy standards like H.264/AVC is driven by AV1's superior compression efficiency, which allows for higher visual quality at significantly lower bitrates. By leveraging this open-source technology, Netflix has enhanced the user experience across a diverse range of devices while simultaneously optimizing global network bandwidth. ### Evolution of AV1 Adoption The journey to 30% adoption began with a strategic rollout across different device ecosystems, balancing software flexibility with hardware requirements. * **Mobile Origins:** The rollout started in 2020 on Android using the "dav1d" software decoder, which was specifically optimized for ARM chipsets to provide better quality for data-conscious mobile users. * **Large Screen Integration:** In 2021, Netflix expanded AV1 to Smart TVs and streaming sticks, working closely with SoC vendors to certify hardware decoders capable of handling 4K and high frame rate (HFR) content. * **Ecosystem Expansion:** Support was extended to web browsers in 2022 and eventually to the Apple ecosystem in 2023 following the introduction of hardware AV1 support in M3 and A17 Pro chips. ### Quantifiable Performance Gains The shift to AV1 has resulted in measurable improvements in video fidelity and streaming stability compared to previous standards. * **Visual Quality:** On average, AV1 streaming sessions achieve VMAF scores that are 4.3 points higher than AVC and 0.9 points higher than HEVC. * **Bandwidth Efficiency:** AV1 sessions require approximately one-third less bandwidth than both AVC and HEVC to maintain the same level of quality. * **Reliability:** The increased efficiency has led to a 45% reduction in buffering interruptions, making high-quality 4K streaming more accessible in regions with limited network infrastructure. ### Live Streaming and Spatial Video Beyond standard video-on-demand, Netflix is utilizing AV1 to power its latest innovations in live broadcasting and immersive media. * **Live Events:** For major live events, such as the Jake Paul vs. Mike Tyson fight, Netflix utilized 10-bit AV1 to provide better resilience against packet loss and lower latency compared to traditional codecs. * **Immersive Content:** AV1 serves as the backbone for spatial video on devices like the Apple Vision Pro, delivering high-bitrate HDR content necessary for a convincing "cinema-grade" experience. As AV1 continues to displace older codecs, the industry is already looking toward the next milestone with the upcoming release of AV2. For developers and hardware manufacturers, the rapid success of AV1 underscores the importance of supporting open-source media standards to meet the increasing consumer demand for high-fidelity, low-latency streaming.

naverOriginal article

Naver TV (opens in new tab)

Processing complex PDF documents remains a significant bottleneck for Large Language Models (LLMs) due to the intricate layouts, nested tables, and visual charts that standard text extractors often fail to capture. To address this, NAVER developed PaLADIN, an LLM-friendly PDF parser designed to transform visual document elements into structured data that models can accurately interpret. By combining specialized vision models with advanced OCR, the system enables high-fidelity document understanding for demanding tasks like analyzing financial reports. ### Challenges in Document Intelligence * Standard PDF parsing often loses the semantic structure of the document, such as the relationship between headers and body text. * Tables and charts pose the greatest difficulty, as numerical values and trends must be extracted without losing the spatial context that defines their meaning. * A "one-size-fits-all" approach to text extraction results in "hallucinations" when LLMs attempt to reconstruct data from fragmented strings. ### The PaLADIN Architecture and Model Integration * **Element Detection:** The system utilizes `Doclayout-Yolo` to identify and categorize document components like text blocks, titles, tables, and figures. * **Table Extraction:** Visual table structures are processed through `nemoretriever-table-structure-v1`, ensuring that cell boundaries and headers are preserved. * **Chart Interpretation:** To convert visual charts into descriptive text or data, the parser employs `google/gemma3-27b-it`, allowing the LLM to "read" visual trends. * **Text Recognition:** For high-accuracy character recognition, particularly in multi-lingual contexts, the pipeline integrates NAVER’s `Papago OCR`. * **Infrastructure:** The architecture leverages `nv-ingest` for optimized throughput and speed, making it suitable for large-scale document processing. ### Evaluation and Real-world Application * **Performance Metrics:** NAVER established a dedicated parsing evaluation set to measure accuracy across diverse document types, focusing on speed and structural integrity. * **AIB Securities Reports:** The parser is currently applied to summarize complex stock market reports, where precision in numerical data is critical. * **LLM-as-a-Judge:** To ensure summary quality, the system uses an automated evaluation framework where a high-performing LLM judges the accuracy of the generated summaries against the parsed source data. For organizations building RAG (Retrieval-Augmented Generation) systems, the transition from basic text extraction to a layout-aware parsing pipeline like PaLADIN is crucial. Future improvements focusing on table cell coordinate precision and more granular chart analysis will further reduce the error rates in automated document processing.