Finetiks AI-Powered Money Management
Niyu Avril
Finetiks AI-Powered Money Management
Helping users forecast expenses up to 98% accuracy with AI-powered Financial Assistant.
A smart, intuitive UI turns AI-driven insights into clear, actionable decisions, giving users confidence and control over their money.
Problem Statement
Users Often:
Don’t track spending consistently
Don’t understand where their money goes
Struggle to set realistic savings goals
Most existing tools either overwhelm users or fail to adapt to their behavior. A common thread: they don’t feel supported in building better financial habits.
"I've been motivated several times to regularly track my finances, but I quickly lose consistency. When I try to continue again, it feels like the records are no longer complete because I missed so much, and that discourages me from continuing."
Context
This project was part of a broader initiative to explore meaningful applications of AI in personal finance.
My Role: Initiated product features and helped design a flow that made it easier for users to record transactions
Collaboration: Worked cross-functionally with UX designers, PMs, CEO, data team, and back-end developers
Objective: Design an AI-powered financial assistant that delivers personalized and actionable insights based on user financial behavior
Relevance
For users: Reduce financial stress by simplifying transaction tracking across platforms, automatically categorizing spending, and offering budget and saving recommendations.
For business: Drive deeper engagement, increase retention, and stand out in a crowded fintech market.
Research Insights
Users often had low financial literacy and lacked consistency in tracking
Spreadsheet users had data but no insights, nothing to help change habits
Budgeting tools felt static, prescriptive, or disconnected from real behavior
Our Solutions
Bank Statement Upload - Simplified onboarding by allowing users to upload e-statements. Generated relevant insights without tedious data entry.
Contextual Recommendation - Developed a behavior-based, conversational recommendation system, users received personalized budgets and realistic savings suggestions.
Learning Loop - The AI assistant adapted over time, forecasted likely expenses, and surfaced smarter suggestions as users engaged.
What Users Said
"I used to set savings goals that exceeded my real capacity, which stressed me out when I couldn’t achieve them. But Finetiks understood my actual saving ability, and I finally succeeded."
"I can now easily track my finances weekly, with the help of the system logging it for me. I no longer worry about missing daily entries."
"My monthly expenses were never consistent, but now I get predictions for future spending, and to my surprise, they’re quite accurate."
Measurable Impact
✅ Achieved 98% forecast accuracy by aligning design with machine learning models in an understandable way
✅ Improved user engagement through personalized recommendations rather than generic calculations
✅ Reduced onboarding drop-off through continuous improvement
What I Learned
The best AI-powered experiences still need to feel human
We must adapt to users’ habits gradually, rather than forcing instant consistency
Clear UI insights were just as critical as machine learning accuracy
If I could revisit the project, I’d invest more in the capability to capture cash transactions and enhance the AI assistant's ability to encourage users.
Stakeholder Feedback
"We need to capture all assets, including credit card debt. A more advanced feature would be the ability to track all assets comprehensively."
"Users should be able to visualize their financial health over time. A trend dashboard or historical comparison could be very impactful."
"The AI assistant needs a clearer onboarding narrative. Users should immediately understand what kind of help they’ll receive and how it evolves over time."
Final Thought
Designing with AI isn’t about showcasing intelligence, it’s about making it useful. My role was to translate complex data into simple, supportive experiences users could trust and act on.
Like this project
Posted May 15, 2025
Helping users forecast expenses up to 98% accuracy with AI-powered Financial Assistant..
Likes
0
Views
0
Timeline
Sep 1, 2022 - May 15, 2025
Clients
FINETIKS
Tokopedia Finance Back Office
TrueMoney EDC (Electronic Data Capture)
2020 - Split! Bill ft. Aisyah Illustration
2020 - Pension Fund Calculator
