AI-First Personal Finance Assistant (Open Banking, Secure MVP)
Budget: €36 – €0 EUR
I am deliberately rebuilding a personal finance web app from scratch, based on concrete learnings from an earlier prototype.
That earlier version validated the problem space (financial stress caused by lack of clarity and guidance), but it also made one thing very clear:
the next version must be significantly leaner, explicitly AI-first, and built on a strong security foundation.
The previous build became too feature-heavy and focused too much on tooling instead of interaction. This rebuild is a conscious reset.
The AI assistant is the product.
All financial data (transactions, categories, subscriptions) exists only to power meaningful, proactive AI conversations.
This MVP is about validation, not scale.
Core goal of the MVP:
To validate whether users trust and engage with a proactive, human, non-judgmental AI assistant that helps them understand and reduce financial stress.
Scope — Phase 1 (Lean, AI-first)
1. Financial data (supporting layer)
- Bank transactions via open banking (EU / PSD2 compliant)
- Account balances
- Very basic categorisation (e.g. groceries, rent, subscriptions, other)
- Simple subscription detection (basic recurring logic is sufficient)
We are planning to use a pay-as-you-go open banking provider (e.g. Ponto or similar), not an expensive upfront enterprise setup.
2. AI Assistant (core of the product)
The AI assistant should:
- Proactively surface insights (maximum 1–2 nudges per week)
- Explain spending patterns in plain, human language
- Never judge, shame or overwhelm the user
- Help users before financial stress escalates, not after
Example interactions:
“Where did my money go last month?”
“Why am I always short at the end of the month?”
“Which subscriptions are quietly costing me money?”
“Can I afford €400 this month?”
“I feel stressed about my finances — help me understand what’s happening.”
The AI should:
- Always explain context before suggesting anything
- Offer one small, realistic next step at a time
- Sometimes reassure instead of optimise
- This is not a budgeting app, dashboard-driven tool, or financial advisor.
- Security & data protection (non-negotiable)
- This product deals with highly sensitive financial data.
Security, privacy and trust are core requirements, not afterthoughts.
Key principles:
- Strong experience with open banking / PSD2
- Secure handling of financial data (encryption, access control, least-privilege)
- Clear separation between raw financial data and AI processing
- GDPR-aware development practices
- No unnecessary data collection
- Security-conscious architecture decisions from day one
- If security is treated as “something we’ll fix later”, this project is not a fit.
- Explicit non-goals (very important)
This MVP will not include:
- Advanced dashboards or charts
- Budget planning systems
- Savings goals or automation
- Employer / HR dashboards
- Compliance-heavy financial advice logic
- If a feature does not directly improve the AI–user interaction, it does not belong in phase 1.
- Long-term collaboration (very important)
I am not looking for a one-off freelancer to “just build something” but to realy make the difference.
I am looking for a long-term technical partner who:
- Wants to help shape the product over time
- Thinks along about architecture and evolution
- Is open to growing with the product beyond the MVP
- Cares about quality, security and user trust
- This MVP is the starting point, not the end.
Tech & working style:
- Open to stack suggestions, but simplicity and security come first
- Clean, minimal UI (transaction list + chat interface)
- AI behaviour will be tested and refined continuously (“boom, boom, boom” iterations)
- Expect frequent discussion around UX, AI tone, and edge cases
Who I am looking for:
- I am looking for a developer (or small team) who:
- Thinks AI-first, not feature-first
- Has experience with open banking / fintech / PSD2 (strong plus)
- Understands security-first development
- Is comfortable with sensitive financial data
- Understands lean MVP thinking and validation
- Is interested in a long-term collaboration, not just a short project
How to apply (important filter):
When applying, please answer the following:
- What experience do you have with open banking / PSD2 integrations?
- How do you typically approach security when working with financial data?
- How would you approach building an AI-first MVP like this?
- What would you deliberately not build in phase 1?
- Are you open to a long-term collaboration beyond the MVP?
Generic or copy-paste proposals will be ignored.
Note on budget & collaboration:
This project is currently bootstrapped, so I work lean with a small consistently budget. I’m explicitly looking for a long-term collaboration with someone who believes in the product and wants to grow it step by step.
That earlier version validated the problem space (financial stress caused by lack of clarity and guidance), but it also made one thing very clear:
the next version must be significantly leaner, explicitly AI-first, and built on a strong security foundation.
The previous build became too feature-heavy and focused too much on tooling instead of interaction. This rebuild is a conscious reset.
The AI assistant is the product.
All financial data (transactions, categories, subscriptions) exists only to power meaningful, proactive AI conversations.
This MVP is about validation, not scale.
Core goal of the MVP:
To validate whether users trust and engage with a proactive, human, non-judgmental AI assistant that helps them understand and reduce financial stress.
Scope — Phase 1 (Lean, AI-first)
1. Financial data (supporting layer)
- Bank transactions via open banking (EU / PSD2 compliant)
- Account balances
- Very basic categorisation (e.g. groceries, rent, subscriptions, other)
- Simple subscription detection (basic recurring logic is sufficient)
We are planning to use a pay-as-you-go open banking provider (e.g. Ponto or similar), not an expensive upfront enterprise setup.
2. AI Assistant (core of the product)
The AI assistant should:
- Proactively surface insights (maximum 1–2 nudges per week)
- Explain spending patterns in plain, human language
- Never judge, shame or overwhelm the user
- Help users before financial stress escalates, not after
Example interactions:
“Where did my money go last month?”
“Why am I always short at the end of the month?”
“Which subscriptions are quietly costing me money?”
“Can I afford €400 this month?”
“I feel stressed about my finances — help me understand what’s happening.”
The AI should:
- Always explain context before suggesting anything
- Offer one small, realistic next step at a time
- Sometimes reassure instead of optimise
- This is not a budgeting app, dashboard-driven tool, or financial advisor.
- Security & data protection (non-negotiable)
- This product deals with highly sensitive financial data.
Security, privacy and trust are core requirements, not afterthoughts.
Key principles:
- Strong experience with open banking / PSD2
- Secure handling of financial data (encryption, access control, least-privilege)
- Clear separation between raw financial data and AI processing
- GDPR-aware development practices
- No unnecessary data collection
- Security-conscious architecture decisions from day one
- If security is treated as “something we’ll fix later”, this project is not a fit.
- Explicit non-goals (very important)
This MVP will not include:
- Advanced dashboards or charts
- Budget planning systems
- Savings goals or automation
- Employer / HR dashboards
- Compliance-heavy financial advice logic
- If a feature does not directly improve the AI–user interaction, it does not belong in phase 1.
- Long-term collaboration (very important)
I am not looking for a one-off freelancer to “just build something” but to realy make the difference.
I am looking for a long-term technical partner who:
- Wants to help shape the product over time
- Thinks along about architecture and evolution
- Is open to growing with the product beyond the MVP
- Cares about quality, security and user trust
- This MVP is the starting point, not the end.
Tech & working style:
- Open to stack suggestions, but simplicity and security come first
- Clean, minimal UI (transaction list + chat interface)
- AI behaviour will be tested and refined continuously (“boom, boom, boom” iterations)
- Expect frequent discussion around UX, AI tone, and edge cases
Who I am looking for:
- I am looking for a developer (or small team) who:
- Thinks AI-first, not feature-first
- Has experience with open banking / fintech / PSD2 (strong plus)
- Understands security-first development
- Is comfortable with sensitive financial data
- Understands lean MVP thinking and validation
- Is interested in a long-term collaboration, not just a short project
How to apply (important filter):
When applying, please answer the following:
- What experience do you have with open banking / PSD2 integrations?
- How do you typically approach security when working with financial data?
- How would you approach building an AI-first MVP like this?
- What would you deliberately not build in phase 1?
- Are you open to a long-term collaboration beyond the MVP?
Generic or copy-paste proposals will be ignored.
Note on budget & collaboration:
This project is currently bootstrapped, so I work lean with a small consistently budget. I’m explicitly looking for a long-term collaboration with someone who believes in the product and wants to grow it step by step.