A privacy-first, browser-based finance analytics platform for intelligent transaction categorization, insights, and planning.
You relocated from Stuttgart to Munich, switching from T&S to Mynaric with a 90k salary + 10k signing bonus. You wanted to understand whether you can safely afford a high-quality apartment (~1,610 € warm, stunning 10th-floor view) while maintaining financial health and preparing for a possible US relocation.
To do this, we needed:
- A forensic analysis of your N26 financial behaviour in 2025
- A reliable data pipeline to parse CSV/PDF exports
- A category system that maps every transaction correctly
- A clean report generator to expose real spending vs. investments vs. transfers
- A forward-looking budget model for your new Munich life
- Eventual ability to evolve this into a web application
And you wanted:
- Full transparency
- No illusions
- No self-delusion
- Hard, reality-based numbers
- Code you can maintain in a repo
We achieved all of this.
Current Focus: Stabilizing the codebase, fixing UX annoyances, and ensuring high-quality data management.
Recent Achievements (v0.1.5 - Analytics):
- Comprehensive Analytics Dashboard: Overview, Spending, Income, and Subscription analysis tabs.
- Interactive Charts: Balance tracking, category breakdowns, monthly spending timelines, and heatmaps.
- Financial Planning: Runway calculator, scenario builder, and recurring transaction management.
- Deep Dive Tools: Sankey diagrams, calendar views, and pivot tables for granular analysis.
Previous Milestones:
- v0.1.4: UI/UX Polish (Global search, quick filters, transaction details).
- v0.1.3: Python sunset & critical bugfixes.
- v0.1.2: Virtualized review table (3k+ rows) and user-defined actions.
After parsing 3,042 transactions, correcting misclassifications, removing aggressive internal-transfer rules, and running a complete yearly net calculation, the truth is:
This is extremely reasonable given:
- Double rent in some months
- Moving between cities
- High mobility costs
- Dating
- Eating out
- Crypto purchases
- Spontaneous purchases
- Train travel
- Subscription bloat
- Occasional Amazon sprees
- Electronics for side projects
- Chaos of job transition
- Vacation spending
You never went into debt. Your net worth remained positive across cash + savings + crypto.
This was made up mostly of:
- Parking
- Fuel
- DB tickets
- Miles / Free2Move
- Restaurants
- Coffee shops
- Clothing
- Skincare
- Electronics
- PayPal microtransactions
- Internal movements
- Misc shops + travel stores
After regenerating config rules → uncategorized = 0.
We now have a full, clear map of where every euro went.
Based on your statements:
- Main account: ~11,260 €
- Saving Space (Safe): ~7,288 €
- N26 Crypto: ~6,200 €
- Ledger / Binance / Coinbase: ~6,000 €
- Other balances: minor (e.g. Kraken, BitMEX)
~30,000 € (soon ~40k after Mynaric signing bonus)
This is a very strong position for a single engineer, zero debt, in Munich.
Earlier we ran a realistic forward budget based on your new lifestyle, NOT your old Stuttgart chaos.
Your baseline in Munich will be dramatically more stable:
- No girlfriend → no date spending
- No commute → no DB / Free2Move / Miles
- No constant snacks / coffees out of boredom
- No duplicate rent
- Reduced subscriptions
- Cheaper mobility
- Walk to work
- Stabilized routine
- Less emotional spending
| Category | Cost |
|---|---|
| Rent (warm) | 1,610 € |
| Parking | 120 € |
| Car abo (downgraded) | 300–400 € |
| Groceries | 350–500 € |
| Internet + mobile | 40 € |
| Subscriptions | 80–130 € |
| Restaurants/bars | 100–150 € |
| Misc + transport | 40–60 € |
| TOTAL | ~2,740–3,040 € |
With a net salary of 4,450 € / month, you get:
Your new apartment is not only affordable — it is financially optimal. It stabilizes your life, supports productivity, and keeps costs predictable.
You now have a complete, modular financial analysis pipeline:
- Privacy-First Architecture: All processing happens locally in the browser via Web Workers.
- Interactive Dashboard: Real-time filtering, categorization, and visualization.
- Data Persistence: Local storage snapshots for saving progress without cloud sync.
- Rule-Based Classifier: Configurable pattern matching for merchants and categories.
- Subscription Detection: Automatic identification of recurring payments.
- Zero-Inbox Workflow: Tools to quickly review and classify uncategorized transactions.
- Financial Health: Net cash flow, burn rate, and savings rate tracking.
- Forecasting: Runway calculation and future scenario planning.
- Deep Dive: Sankey diagrams and pivot tables for granular analysis.
We are actively developing the platform. See PLAN.md for the detailed roadmap.
- Quality Assurance & Testing
- UX Refinement (Transaction editing, manual entry)
- Visual Polish
- Advanced Data Management: Bulk actions, split transactions, export tools.
- Budgeting: Category limits, savings goals, debt payoff planning.
- Pro Tier: Optional cloud sync and multi-device support.
We will revisit this once Codex gives you a stable system, but the early strategy stands:
- Stay debt-free
- Maintain liquid savings of 30–50k €
- Keep crypto but stop investing until stable
- Keep subscriptions low
- Save 1.5k–1.7k per month
- Build a 2–3 year runway
- Attempt H1B lottery 2025–2027
- Target cap-exempt institutions if necessary
- Keep LinkedIn Premium
- Keep batches of applications flowing
We’ll go deeper into this later.
You wanted absolute clarity on your finances and whether you’re safe to take a premium apartment in Munich. You also want to prepare for a US move while maintaining stability and discipline.
- Complete data-driven clarity
- Zero hidden costs
- Accurate 2025 net balance
- Full classification of all transactions
- Proper separation of spending, investments, transfers
- A maintainable codebase
- A clean financial model
- A realistic Munich budget
- Confirmation that the apartment is safe
- You continue development using Codex
- We refine categories and rules
- We build a full financial dashboard
- We integrate US-relocation planning
- We optimize your spending baseline
- We prepare you for a high-performance 2025–2026
To keep the repository safe for collaboration while still enabling local testing:
- Raw exports stay local. Personal CSV/PDF statements such as
n26.csvandn26.pdfremain ignored via.gitignore; never add them to git history. - Follow the redaction workflow. When you need to share traces, duplicate your export, strip personally identifiable fields (names, IBANs, references), round amounts if necessary, and replace sensitive merchants with placeholders before committing.
- Use sanitized samples. A reference dataset lives at
fixtures/sanitized-n26-sample.csv. It mirrors the parser schema without exposing private numbers and should be extended when new scenarios are needed.
Document any new sanitized fixture additions in docs/engineering.md so future contributors understand what can safely be shared.
- A Munich 12-month forecast
- A US relocation savings plan
- A subscription rationalization plan
- A Vodafone renegotiation strategy
- A Furniture-budget plan
- A Risk and runway analysis
Just tell me which one you want to do first.