Developer builds AI app that speaks to anxious dogs in their owner's voice

·1 views

A developer has created a web app called Stay that uses AI to comfort dogs left home alone by speaking to them in their owner's synthesized voice. The app listens for distress sounds and, crucially, waits for the dog to go quiet before responding — a deliberate design choice to avoid reinforcing the barking. Unlike a 2021 Finnish app that played looped voice recordings and risked conditioning dogs to associate the sound with abandonment, Stay generates fresh phrases each time to prevent that learned fear response. The AI-generated speech follows strict constraints, including short sentence lengths, no questions, and no references to leaving or returning, to keep the tone calm and neutral. A demo version is publicly accessible without any account or API key, and the project's code is available on GitHub.

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)

How to Set Up and Run Your First Load Test Using Apache JMeter

Apache JMeter is a widely used open-source tool for load, stress, and performance testing of applications. Getting started requires Java JDK 11 or higher, after which JMeter can be downloaded from the official Apache site and launched from its bin directory. Users can build a basic test plan through the GUI by configuring a Thread Group, HTTP Request Defaults, and an HTTP Sampler to simulate virtual users. For accurate heavy load testing, JMeter recommends running tests in non-GUI command-line mode, which also generates an interactive HTML report of results. An optional Plugins Manager can be installed to extend JMeter's capabilities with additional listeners, graph generators, and custom samplers.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

8 React Native Folder Structures Explained: How to Pick the Right One

A React Native team lead has outlined eight real-world project folder structures, explaining the trade-offs each one addresses for apps of different sizes and team scales. The guide covers approaches ranging from a flat single-directory layout suited for demos and hackathons to feature-based and layer-based structures used in production apps. It highlights how feature-based architecture maps cleanly to team ownership and reduces merge conflicts, but warns that cross-feature imports can silently undermine isolation without enforced lint rules. The article also distinguishes when simpler structures suffice versus when more complex patterns like domain-driven or micro-frontend layouts are warranted. The goal is to help developers reason about architectural trade-offs rather than defaulting to over-engineered setups for early-stage projects.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Snowflake's MATCH_RECOGNIZE Used to Detect Hidden Pain Patterns in Dogs

A developer has built a data project called TELLTAIL that uses Snowflake's MATCH_RECOGNIZE feature to identify early signs of chronic pain in dogs by analyzing movement sequences rather than simple step counts or averages. The project was submitted for a Weekend Challenge coinciding with International Dog Day on August 26. The motivation stems from research showing that 38% of dogs have osteoarthritis, yet only around 2.5% are diagnosed annually, largely because dogs instinctively conceal signs of pain. Traditional consumer fitness trackers compress movement data into scalar values, which erases the sequential patterns that carry the actual clinical signal. The tool aims to make findings explainable down to the individual second and logs each published claim to a public ledger for transparency and accountability.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How to Run Playwright and Cucumber UI Tests on LambdaTest via Azure DevOps

Developers can integrate Playwright and Cucumber test suites with LambdaTest Cloud by configuring an Azure DevOps pipeline connected to a GitHub repository. The setup requires an active Azure DevOps account, a LambdaTest account with credentials, and a GitHub repo containing the automation code. LambdaTest credentials are stored securely as a variable group in Azure DevOps Pipelines Library and injected into the pipeline at runtime. An azure-pipelines.yml file placed at the repository root defines the build stages, installs Node.js dependencies, and triggers UI smoke tests on LambdaTest's remote browser infrastructure. Once configured, every push to the main branch automatically runs the pipeline, with JUnit test results published in Azure DevOps and video logs available on the LambdaTest dashboard.

0 comments Read more at DEV Community