Solo dev builds AI-powered movie tracker app for couples to manage shared watchlists

·1 views

A Netherlands-based full-stack developer spent a year building 'I Like Movies', an Android app launched on Google Play in summer 2025, designed to help households manage a shared watchlist and watched history. The app was created to solve the real-world problem of two people struggling to agree on what to watch, with changes syncing instantly across all household devices. The backend runs on Go, GraphQL, and Postgres, while the frontend uses React Native with Expo, a deliberately conservative stack chosen for solo maintainability. A standout feature is an AI chat assistant that recommends films in plain language, using tool-calling to verify every suggestion against TMDB metadata, the user's watch history, and local streaming availability. Over-the-air updates via Expo allow the developer to push JavaScript-level bug fixes to all users within minutes, without waiting for app store review.

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)

Loupe brings on-device network and storage debugging to React Native apps

A developer has released Loupe, an open-source debug overlay tool for React Native that displays network activity, logs, storage data, and deep links directly within the app on a real device. The tool requires just a single import line in index.js and needs no desktop client, USB cable, or additional configuration. Loupe is automatically excluded from production builds by design, as all debug code sits behind the __DEV__ flag and is stripped out by Metro during release compilation. It offers first-class GraphQL support by labeling requests with operation names and flagging HTTP 200 responses that contain errors, addressing a common blind spot in standard network inspectors. Sensitive storage values such as auth tokens remain masked by default, with copy functionality disabled until the user explicitly reveals them.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Argues BlocSignal Cuts Flutter State Management Complexity Over Riverpod

A Flutter developer and former Riverpod advocate has published a detailed critique arguing that Riverpod's evolution across versions has accumulated significant cognitive overhead for developers. The author contends that modern Riverpod requires mandatory code generation, obscure base class extensions, and constant build_runner processes that slow down development workflows. While acknowledging Riverpod creator Rémi Rousselet's foundational contributions to Flutter architecture, the author argues that tooling aids like linters and IDE plugins only mask deeper architectural complexity rather than resolving it. The piece promotes BlocSignal, a combination of the BLoC/Cubit pattern with Signals-based reactivity, as a simpler alternative that eliminates much of that mental overhead. The author frames the switch not as gaining new features, but as shedding unnecessary complexity accumulated over years of framework evolution.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Builds Subscription Management System Using State Machine and AI Tools

A developer has shared a subscription management system built using a state machine concept, with a stack comprising C#/.NET 10, Entity Framework Core, PostgreSQL, ReactJS 19 with TypeScript, and Docker. The project handles subscription plans, subscriber management, and plan transitions such as moving from trial to paid tiers. Payment platform integration was intentionally excluded at this stage, as the goal is to validate the core workflow first. The developer noted that AI-assisted development compressed what might take a typical programmer four weeks into roughly four hours, though this speed introduced technical debt through duplicated and redundant code. The author emphasizes that thorough upfront planning remains critical to reducing production-phase issues, regardless of how fast AI tools can generate code.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How to Detect Hallucinations in RAG Systems Using RAGAS and Other Tools

In Retrieval-Augmented Generation (RAG) systems, hallucination occurs when a large language model generates a response from its own training data rather than from the provided document context. One detection approach involves comparing the vector embeddings of the model's response with those of the retrieved context to measure semantic closeness. LettuceDetect, a BERT-based model, can identify potential hallucinations at the character level by analyzing responses word by word. RAGAS, an evaluation framework, offers multiple metrics — including Faithfulness, Answer Relevancy, and Context Recall — to systematically assess RAG output quality and can be integrated into CI/CD pipelines for automated testing. DeepEval is another available tool that can similarly be used to evaluate RAG systems and flag hallucinated responses.

0 comments Read more at DEV Community