Labour & error savings
staff × hrs/wk × 50 weeks × wage × auto% for hours, plus 70% reduction on error rate × cost/error. Both in MYR, fully-loaded.
Move the inputs to model annual labour savings, token spend, platform fees, and one-time implementation cost. Calibrated for Malaysian salary benchmarks and 2026 model pricing.
staff × hrs/wk × 50 weeks × wage × auto% for hours, plus 70% reduction on error rate × cost/error. Both in MYR, fully-loaded.
requests/day × 250 working days × (in_tokens × in_rate + out_tokens × out_rate). Tier rates are 2026 public model pricing converted to MYR — Fast (Haiku-class), Balanced (Sonnet-class), Advanced (Opus-class).
Teragrid Ai PaaS: RM 18,000 base + RM 6,000 × workers per year. Implementation is one-time and amortised across the 3-year ROI calculation.