Reserva applies algorithmic optimization to project-based income, modeling cash flow gaps and surplus periods so freelancers in Germany can allocate capital with measurable risk boundaries instead of guesswork.
The model does not predict the market. It predicts your exposure to it, based on the rhythm of your own project income.
Freelancers rarely have a single monthly figure to optimize around. Reserva ingests historical transaction patterns and upcoming contract data, then runs a rolling forecast that flags periods where capital is likely to sit idle or fall short.
Financial data from independent professionals is treated with the same operational discipline expected in regulated institutions, without adding friction to daily use.
All stored data is protected end-to-end.
Practices are structured to reflect current German and EU requirements.
All client data is processed and stored on servers located within Germany. No transaction data is used to train models shared across unrelated accounts, and deletion requests are executed on a fixed schedule in line with DSGVO obligations.
Both run continuously in the background and adjust their output as new transaction data arrives.
The system monitors incoming and outgoing flows against a rolling 90-day baseline and flags deviations that could indicate an upcoming liquidity gap, before it becomes a cash flow problem.
Idle capital above a configurable reserve threshold is flagged with allocation options ranked by liquidity need and time horizon, rather than a single blanket recommendation.
A rolling timeline lets you drag projected contract end dates and immediately see the recalculated reserve and exposure figures, without submitting a new report.
Transparency on method matters more than a polished dashboard when the output affects real capital decisions.
Reserva combines time-series forecasting with a constrained optimization layer. The forecasting component estimates likely income and expense ranges; the optimization layer then allocates capital across reserve, short-term, and growth buckets within limits you define.
Inputs come from your connected accounts or uploaded statements, combined with publicly available macroeconomic indicators such as interest rate benchmarks. No third-party behavioral data or scraped personal profiles are used in the model.
Forecast accuracy is back-tested against historical income patterns from each account before recommendations are activated, and the model is re-validated whenever a material shift in income structure is detected.
Direct answers, without marketing language, for professionals evaluating Reserva against internal compliance checklists.