The models behind the numbers.

Problem: A PE sponsor needed an independent, defensible valuation before committing capital to a buyout with a complex structure: senior debt tranches, PIK loan notes, management rollover, and a sweet-equity pool. The sponsor also had to know whether the entry multiple was justified across exit horizons.

Approach: Built an integrated LBO and DCF model from a single assumptions tab: full sources-and-uses, a diluted cap table with management rollover and sweet equity, a multi-tranche debt and PIK interest schedule, a returns waterfall by stakeholder class, and a bottom-up WACC (relevered beta, CAPM) with both perpetuity-growth and exit-multiple terminal values; valuation triangulated across LBO and DCF.

Result: A defensible valuation range of ~7.2x–10.8x EV/EBITDA, converging near the ~8.2x entry multiple and confirming the price was supportable. Returns modeling showed sponsor IRRs scaling from ~5% on an early exit to ~23% by 2030 (3.0x MOIC), with management reaching ~4.5x, giving leadership a stakeholder-by-stakeholder basis to negotiate and structure the deal.

Problem: Revenue had nearly doubled in two years, but demand planning hadn't scaled: one planner ran a single blended forecast across a start-up beverage line, a scaling snack line, and a mature core-SKU line. The result: simultaneous stockouts and excess inventory, no reconciliation of forecast to board plan, unmodeled promotions, no accuracy KPIs, and key-person risk.

Approach: Stood up a documented, repeatable demand-planning discipline: separate forecast logic per line, reconciled to a single board-level view, with variances traced to specific drivers and accuracy tracked against a formal KPI rather than explained after the fact.

Result: Forecast accuracy reached ~8.7% MAPE with near-zero bias inside the pilot window; the mature line held a 100% hit-rate within ±5% of forecast every month. The function moved from a fragile, single-person spreadsheet to a scalable process resilient to key-person risk.

Problem: A sponsor needed an investment-grade model to test viability before committing capital and approaching lenders: USD-priced capital goods against KES-linked revenue, uncertain financing terms, and an unverified headline cost figure left returns, debt capacity, and downside risk all unknown.

Approach: Built a 10-year integrated model with a four-scenario engine (downside/base/sponsor/upside) driving occupancy ramp and ADR across all tabs: a bottom-up CAPEX with freight/install multipliers and contingency, a RevPAR-driven revenue build, an amortizing debt schedule with DSCR, and two-way IRR sensitivity grids on the key value drivers.

Result: A defensible return range across scenarios (base-case unlevered IRR ~18%, levered equity IRR ~21%, MOIC ~6x) and a stabilized EBITDA margin near 39%. Surfaced two decision-critical findings: minimum DSCR fell below 1.0x during ramp-up (flagging need for an interest reserve), and bottom-up CAPEX exceeded the sponsor's stated budget, exposing a Phase 1 under-budgeting gap.