PocketPilot - Your Smart Financial Co-Pilot

Abstract

In this work, we propose PocketPilot: a smart financial assistant, designed to facilitate personal expense tracking, use artificial intelligence for intelligent decision making, and demonstrate its implementation with an active learning pipeline. The application is built on many emerging technologies, including OCR for taking pictures of receipts, NLP for intelligent category selection, and voice processing for people-friendly data entry. Our tool plays a predictive role in achieving financial goal and utilizes machine learning to suggest modifications in its predictions. Expense capture from receipts, SMS alerts, or voice input: thereafter, the system shall automatically classify the expenses to the most likely categories such as Rent, food, travel, medical based on AI and machine learning. The system shall predict future spending using machine learning, thus enabling people to budget for things which are budgeted for such health emergencies. The feature shall also allow the customers to receive recommendations or financial advice based on the spending trend. The project builds upon limitations identified in existing research, particularly the IEEE paper ‘‘AI-Driven Financial Insights for Personal Budget Planning,” by introducing a more proactive and intelligent approach to financial wellness. This paper outlines the system’s architecture, methodologies, and implementation roadmap, demonstrating its potential as a student-led innovation in AI-powered expense management.

Introduction

The text presents PocketPilot, an AI-powered personal finance tool designed to overcome the limitations of traditional expense trackers, which rely heavily on manual data entry and static categorization. PocketPilot automates expense logging, intelligently categorizes transactions using machine learning and natural language processing, and provides predictive analytics to help users plan future spending and savings. It captures data from multiple sources—including receipts (via OCR), SMS alerts, and voice commands—making the system more flexible and user-friendly.

Inspired by the IEEE paper “AI-Driven Financial Insights for Personal Budget Planning,” PocketPilot builds on earlier forecasting models like ARIMA and Random Forest but addresses their shortcomings, such as poor handling of irregular spending patterns and lack of voice interaction or dynamic budgeting. The literature review emphasizes the need for a more adaptive and intelligent system that guides users proactively rather than merely tracking expenses.

The methodology outlines PocketPilot’s technical architecture: a React.js PWA frontend; a Node.js and Express.js backend using REST APIs and JWT authentication; and a MongoDB Atlas database designed to store user data, expenses, categories, and goals. Several AI modules power the system, including Mindee for OCR receipt extraction, NLP tools for smart categorization, a custom time-series model for forecasting, goal-oriented budgeting, and a voice assistant built with the Web Speech API.

Implementation followed a modular, test-driven approach. Authentication was set up using JWT and bcrypt, while expense logging supported both manual and automated methods. NLP-based categorization assigns accurate labels even when wording varies. Goal tracking offers real-time progress visualizations and alerts, and predictive insights forecast spending trends by analyzing historical patterns and anomalies. Together, these features position PocketPilot as an intelligent, proactive financial co-pilot designed to improve user financial wellness through automation, personalization, and accessibility.

Conclusion

PocketPilot is an attempt in bringing modern web technologies and light-weight?AI modules to deal with current challenges like personal expense management. By combining OCR-based receipt scanning, NLP-driven categorization, predictive analytics and voice interface Dash is a multi-modal intelligent?platform for managing and optimizing finances. Modularity, scalability and end-user?design dominate the architecture. Each component from the React. Js frontend to the Node. Js?as backend and MongoDB as database were chosen for easy integration and maintainability. Using Mindee for OCR, natural.js and compromise.js for NLP, and custom time-series algorithms for forecasting isn’t just about showing off tech it’s about finding that sweet spot between solid performance and keeping things accessible for everyone. But PocketPilot isn’t only about the nuts and bolts. It’s right in the middle of the bigger conversation on AI and financial wellness. The way it automates key tasks actually makes life easier less mental juggling, better budgeting, and smarter decisions all around. You can toss in data in different formats, and PocketPilot still figures out your spending habits and helps you work toward your financial goals. That’s what makes it more than just a proof of concept it’s something real people can actually use.

References

[1] Sharma, R., & Mehta, A. (2022). AI-Driven Financial Insights for Personal Budget Planning: A Smart Approach to Future Expense Prediction. IEEE Xplore. [2] Mindee. (n.d.). OCR API for Receipts and Invoices. Retrieved from https://www.mindee.com [3] Natural.js. . (n.d.). Natural Language Processing for Node.js. Retrieved from https://github.com/NaturalNode/natural [4] Compromise.js. . (n.d.). Fast NLP for JavaScript. Retrieved from https://github.com/spencermountain/compromise [5] Web Speech API. (n.d.). Speech Recognition in Browsers. Retrieved from https://developer.mozilla.org/en-US/docs/Web/API/Web_Speech_API [6] MongoDB Atlas. (n.d.). Cloud Database Service. Retrieved from https://www.mongodb.com/cloud/atlas [7] Vercel. (n.d.). Frontend Deployment Platform for React and Next.js. Retrieved from https://vercel.com [8] Render. (n.d.). Cloud Hosting for Web Services and APIs. Retrieved from https://render.com

Copyright

Copyright © 2025 Dr. Pradeep, Vanshika Rajpal, Manya Garg, Sourabh Goyal , Rishabh Vashist . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.