ArthNiti
Predict. Plan. Prosper.
The problem ArthNiti solves
The Problem It Solves
Most everyday users struggle with managing their personal finances not because they lack income, but because they lack clarity, guidance, and simple financial tools tailored to their real-life behavior.
Existing financial apps are either:
Too complex (designed for investors, not beginners)
Too generic (static budgeting trackers with no personalization)
Not explainable (users don’t understand why recommendations are given)
Dependent on complete financial data (which many users cannot accurately provide)
As a result, users face:
Poor budgeting and uncontrolled spending
Low and inconsistent savings habits
Lack of emergency financial planning
Unclear goal planning (car, education, business, etc.)
No visibility into their financial future
Anxiety and confusion about financial decisions
Additionally, most tools do not account for:
Incomplete financial inputs
Behavioral spending patterns
Risk awareness for unexpected expenses
Low financial literacy users
This creates a major gap between financial awareness and actionable decision-making for the average person.
What People Can Use It For (Real-World Use Cases)
💳 Smarter Budgeting & Spend Awareness
Users can understand where their money is going through AI-powered spend analysis, category breakdowns, and overspending alerts — even with incomplete or estimated data.
💰 Personalized Savings Optimization
The platform analyzes a user’s income, expenses, and behavior to:
Calculate savings rate
Suggest realistic savings targets
Recommend smart budget reallocation
Identify unnecessary spending leaks
This makes saving more structured and achievable.
🎯 Goal-Based Financial Planning
Users can set real-life goals such as:
Emergency fund
Education
Business
Major purchases
The system then provides:
Feasibility score
Required monthly savings
Timeline adjustment suggestions
Risk warnings
🚀 AI Financial Future Simulation (Key Innovation)
Users can simulate their financial future using interactive “what-if” scenarios:
Current spending behavior
Optimized AI plan
Unexpected expense scenarios
This helps users visualize the long-term impact of their financial habits, making abstract financial planning more understandable and actionable.
🧠 Financial Guidance for Low Financial Literacy Users
Unlike traditional financial tools, the platform:
Uses simple language (no complex financial jargon)
Provides explainable AI insights
Works with approximate and incomplete inputs
Focuses on everyday financial decisions, not investments
This makes it accessible to students, salaried individuals, and first-time financial planners.
🛡️ Risk-Aware Financial Safety Insights
The system evaluates:
Emergency fund readiness
Financial stability
Risk vulnerability to sudden expenses
And alerts users with clear, explainable recommendations to improve financial safety.
How It Makes Existing Tasks Easier & Safer
Easier Than Traditional Finance Apps
No need for exact financial documents
Short intelligent onboarding instead of long forms
Automated insights instead of manual calculations
Unified dashboard instead of multiple tools (budget + savings + goals + simulation)
Safer & More Responsible Than Typical Fintech Tools
Avoids regulated investment advice
Provides explainable, transparent recommendations
Includes consent, privacy, and data security considerations
Focuses on financial planning and behavior, not risky speculation
In One Line
Our platform acts as an AI Financial Copilot for everyday users, transforming raw financial behavior into clear insights, future projections, and actionable recommendations — making personal finance simpler, more explainable, and more accessible to everyone.
Challenges we ran into
Challenges I Ran Into
- Designing for Low Financial Literacy Without Oversimplifying
One of the biggest challenges was balancing simplicity and intelligence.
Our problem statement specifically targets everyday users with limited financial literacy, which meant we could not use complex financial jargon, long onboarding forms, or technical inputs like exact expense ledgers and asset allocations.
Initially, our questionnaire became too long and data-heavy because we were trying to capture income, expenses, risk profile, savings, and goals in detail. This created poor UX and friction during onboarding.
How we solved it:
We redesigned the system into:
A condensed core profiling questionnaire (high-yield questions)
Modular micro-quizzes inside each feature (Spend, Savings, Goals)
This allowed us to collect rich behavioral data while keeping the user experience fast, intuitive, and beginner-friendly.
- Handling Incomplete and Approximate Financial Data
Another major hurdle was that real users often do not know their exact financial numbers (monthly expenses, savings %, etc.), but our AI modules (dashboard, simulator, savings engine) required structured inputs.
If we relied only on precise numeric inputs, the system would break or give inaccurate outputs.
How we solved it:
We shifted from exact inputs to:
Range-based questions (income bands, spend ratios)
Behavioral indicators (budgeting habits, expense tracking)
“Not sure” options with fallback assumptions
We then built a rule-based estimation layer that converts approximate answers into usable financial parameters for projections and insights.
- Avoiding Regulated Investment Advice While Giving Allocation Suggestions
Our Savings Optimization module initially included direct suggestions like allocating money into stocks, mutual funds, or debt instruments. This raised a compliance concern because the problem statement explicitly restricts regulated investment advice.
How we solved it:
We reframed the feature as:
Educational savings allocation guidance (Safety, Growth, Liquidity buckets)
Explainable AI recommendations instead of prescriptive financial advice
Clear disclaimers across the dashboard and simulator
This ensured the system remained responsible, compliant, and aligned with the problem constraints.
- Building an Explainable AI Financial Simulator (USP Feature)
The Financial Future Simulator was technically challenging because we needed:
Multi-scenario projections (Current, Optimized, Risk)
Real-time recalculations
Explainable outputs (not black-box AI)
A fully ML-based model was not feasible within hackathon time constraints.
How we solved it:
We implemented a hybrid approach:
Rule-based financial projection logic (income, expenses, savings loops)
Scenario modeling using adjustable variables (sliders)
Dynamic graphs with explainable AI text generation
This allowed us to create an interactive, intelligent simulator that is both lightweight and demo-ready.
- Integrating Spend Analysis with Flexible Data Sources
Our Smart Spend Analysis engine was designed to support email statement parsing (via n8n + RAG), but integration reliability during development was a concern, especially under hackathon time limits.
How we solved it:
We built a fallback-first architecture:
Primary: Email statement parsing (AI categorization)
Secondary: CSV upload
Fallback: Manual expense input
This ensured the system remains functional even if external integrations fail, improving robustness and demo reliability.
- Feature Scope vs Hackathon Time Constraint
We initially ideated a very large feature set including full financial profiling, AI advisor, simulator, savings allocation, retirement planning, and spend intelligence. The challenge was preventing feature bloat while still delivering a polished MVP.
How we solved it:
We introduced strict feature prioritization:
Core MVP: Dashboard + Profiling + Simulator (USP)
Secondary: Savings & Spend intelligence
Future scope: Advanced integrations and predictive modeling
This helped us focus on a working, impactful prototype rather than an incomplete overbuilt system.
- Making the AI Feel Truly “Intelligent” in the Demo
A key hurdle was ensuring the platform didn’t feel like just a static dashboard with charts. Judges expect personalization and reasoning from an AI-driven system.
How we solved it:
We embedded:
Persona classification (e.g., Impulsive Spender, Balanced Planner)
Explainable insight cards
Scenario-based AI explanations
Context-aware recommendations across modules
This significantly improved perceived intelligence and user trust in the system.
Final Reflection
Overall, the biggest challenge was not just technical implementation, but designing a system that is accurate, explainable, compliant, and usable by non-expert users all within hackathon constraints.
We overcame this by prioritizing modular AI logic, explainability, fallback mechanisms, and user-centric UX design
Technologies used