How Cuentopia Built GDPR-Compliant Encryption for Children's Emotional Data

·1 views

Cuentopia, a Spanish app that generates personalized stories for children, discovered it was storing sensitive emotional profiles of minors in its database, prompting a serious review of data protection practices. The team implemented envelope encryption using a per-user key system, with a master key stored securely in Google Cloud KMS that never leaves the service. The architecture was designed to encrypt sensitive fields before they reach the database, reducing exposure risk from breaches, unauthorized admin access, or mishandled backups. An external adversarial audit identified gaps that led to further hardening, including crypto-shredding for data deletion. The entire system was built and tested before any real user data was onboarded, in compliance with GDPR and Spain's LOPD-GDD regulations governing specially protected data categories.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Developer Breaks Free From Tutorial Loop by Committing to Building Projects

A self-taught developer shares how they fell into a cycle of endlessly consuming tutorials and courses without actually building anything meaningful. The pattern felt productive but left them unable to work independently from a blank editor when real projects demanded it. After stepping away from coding and returning with a fresh perspective, they completed a Frontend Mentor Testimonials Grid challenge as a deliberate reset point. The experience reinforced that practical building exposes problems no tutorial can fully prepare a developer for. They now advocate starting projects first and learning along the way, rather than waiting to feel fully ready before writing a line of code.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

One Build Script Publishes a Game Ontology as HTML, Graph, Search, and AI Tools

A developer behind 'The Genome of Games' has built a single Node.js script that publishes 1,180 records across four distinct surfaces — static HTML pages, an interactive graph, a search index, and a Model Context Protocol (MCP) server — in under half a second with no npm dependencies. The knowledge base covers 168 game mechanics, 618 games, and 394 companies spanning 1962 to 2025, sourced from six hand-edited JSON files totalling around 1.3 MB. The build expands that source data 12.7× into over 16 MB of generated output, all of which can be deleted and fully regenerated on the next run. A key architectural decision was to have the MCP server statically import a pre-built index rather than query a live API, ensuring that AI agents and web crawlers always derive answers from the same single source of truth. The approach is presented as a reusable pattern for any structured knowledge base that must serve both search engines and AI models consistently.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Builds 30ms Web Scraper API Optimised for AI Agent Frameworks

A developer has released Smart Scraper M2M, a lightweight web scraping API designed to address performance bottlenecks in AI agent frameworks such as CrewAI and LangChain. The tool returns clean, structured JSON in approximately 30 milliseconds by stripping out irrelevant HTML and CSS before data reaches a language model. This approach aims to reduce context window bloat and lower LLM token costs in machine-to-machine workflows. The project is open source and available on GitHub, where the developer is actively seeking community feedback and contributions.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How to Build a Rate Limiter: Design Principles and Common Pitfalls Explained

Developer Maneshwar, creator of the open-source AI code review tool git-lrc, has published a technical guide on designing production-ready rate limiters. A rate limiter controls how many requests a client can make within a set time window, protecting systems from overload and preventing any single user from consuming disproportionate resources. The guide outlines key requirements including configurable per-user limits, honest HTTP 429 rejections with informative headers, and sub-3ms latency at P95. It walks through the pitfalls of naive approaches such as in-database counters and per-server in-memory storage, both of which break down under real-world, multi-server conditions. Redis is recommended as the shared, memory-fast backbone for counters, using atomic operations and automatic key expiry to keep limits consistent across all servers.

0 comments Read more at DEV Community