What I'm doing now
Business analysis, then the software the analysis calls for
Since July 2026 I've been consulting independently for small and mid-sized
businesses in Jalisco: going into the real operation, mapping the process as it
actually runs, crossing the numbers, and then building what the findings point at.
On a recent engagement with a medical-apparel manufacturer, I consolidated raw
sales, invoicing and accounting exports into a single analytical view. The
headline finding wasn't in any of their reports:
99days of DSO, measured — never previously calculated
~50%of invoiced revenue sitting uncollected
330data anomalies across disconnected systems
94spellings of one customer's name, making per-client margin uncomputable
The finding I'm proudest of took reading formulas rather than numbers: a
current-year P&L that reconciled perfectly against its own parts while its
revenue row summed last year's workbook. It surfaced only because the
analysis crossed the books against a source the client doesn't type by hand —
their stamped tax receipts. The readers that made that possible are open source
as mxbooks.
I also mapped the full order-to-invoice flow in per-role swimlanes, reconciled
real per-product margin against the published price list, and scoped a
mobile-first inventory system built around how the operations lead actually works
on the floor.