{
  "id": 14202419,
  "title": "I built a budgeting app where the AI never does the math",
  "url": "https://urgent.news/2026/10/13/i-built-a-budgeting-app-where-the-ai-never-does-the-math",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-13T13:06:34.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/ostyl51/i-built-a-budgeting-app-where-the-ai-never-does-the-math-3j6e"
  },
  "original_language": "en",
  "account": "Most budgeting apps merely show what you've spent, but Budget Lover addresses a different question: how much you can spend without regret on the 28th. This Android app, built by a Computer Science student, uses AI to provide a daily safe-to-spend amount and an AI advisor to answer follow-up questions. The app is still in development and currently tested on an emulator using demo data.\n\nThe core of Budget Lover's functionality lies in a Dart engine that calculates your \"safe to spend\" amount by subtracting your upcoming bills, subscriptions, planned payments, savings contributions, and a safety buffer from your account balance. It then determines a daily allowance by dividing this safe amount by the days until your next payday. This approach allows for thorough testing and the ability to handle various edge cases, such as month-end changes and moving paydays.\n\nTo create an AI advisor, the app performs calculations locally to generate relevant facts, then sends these facts to Gemini to provide a plain language explanation of the results. This separation of logic and explanation ensures that the AI only focuses on tone and explanation, not actual accounting. Additionally, the app includes a \"what if?\" simulator that compares your current data with hypothetical scenarios—like spending more or cancelling subscriptions—to demonstrate the impact of potential changes.\n\nSeveral features of the app were developed to enhance its functionality. Receipt scanning uses a lightweight Gemini model to extract and validate transaction details, subscriptions categorize recurring payments for easy budgeting, and a trip feature calculates costs per trip, day, and person while providing live currency conversions. The monthly wrap-up presents a story-style summary that can be exported to PDF. The app also offers XP and badge rewards for maintaining good budgeting habits and supports 15 languages, including right-to-left scripts like Arabic.\n\nThroughout the development process, the author learned the importance of separating the app's logic from its user interface. This separation not only simplifies the development process but also allows for more efficient testing, particularly of the more mundane aspects of the app. The author recommends keeping calculations simple and using AI for language-related tasks rather than numerical ones. Testing the app thoroughly is crucial, as mathematical errors are less common than issues related to date handling or other technical aspects.",
  "summary": "Most budgeting apps answer one question: what did I spend? The question I actually ask myself is different: how much can I spend today without regretting it on the 28th? So I built Budget Lover , an Android app (Flutter + Firebase + Gemini) that answers that second question with one number, a daily \"safe to spend\" amount, and an AI advisor you can ask follow-up questions. I'm a Computer Science…",
  "key_points": [
    "Budget Lover app calculates safe-to-spend amount using Dart engine",
    "AI advisor generates plain language explanations for budgeting results",
    "App includes receipt scanning, subscriptions categorization, and trip cost calculations"
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}