My goal was specific: reach a S$14,000 cash buffer within 9 months using AI forecasting to guide spending and savings decisions. Starting balance was S$2,200.
Month 1 to 3: The tool was too optimistic
The AI projected I could save S$1,600 per month based on my average income over the prior year. It did not account for a seasonal dip in March, when 2 of my regular clients paused projects. Actual savings in that window came to S$940 per month.
The model recalibrated after 3 months of new data and reduced its monthly savings projection to S$1,150, which was closer to reality.
Months 4 to 9: More useful signals
Once the forecast stabilized, I found the weekly spending breakdowns genuinely helpful. The system identified that my SaaS subscriptions had grown to S$480 per month without me noticing. Cutting 4 unused tools freed up S$210 monthly.
Final result
After 9 months I reached S$11,600, not S$14,000. The AI gave me structure and spotted patterns I missed manually. It did not compensate for income volatility or early miscalibration. Treat the first quarter of data as a warm-up, not a baseline.