Automotive · Procurement & Spend Control
Approvals lived in inboxes. Spend surfaced at the invoice.
The problem. Across a multi-site automotive operation, purchase approvals ran through disconnected email threads, spreadsheets, and department-level systems. Finance learned about commitments when the invoice arrived — after the decision was made and the money was effectively spent. No one could answer "who approved this, and against what budget?" without an archaeology project.
The intervention. We mapped the approval capability across departments, designed a single approval architecture with routing by amount, department, vendor, and category — then deployed RFPO and integrated it with the existing finance systems rather than replacing them.
What changed.
- Spend became visible to finance before commitment, not after the invoice
- Approval cycles compressed from days of email-chasing to routed, tracked decisions
- Every approval carries an immutable audit trail a controller and an auditor can both read
- The repeated problem became a product — RFPO is now our flagship
Professional Services Margin
Every statement of work was written from scratch — by the most expensive person available.
The problem. Scope documents took senior practitioners days to draft, each author worded them differently, and the protective assumptions were the first casualty of deadline pressure. The cost showed up later, as change orders and disputes born from ambiguity nobody caught.
The intervention. We built an approved clause and assumption library from the firm's best prior work, put a structured intake in front of it, and used AI-assisted generation — the pattern that became SOW Generator — so experts review and sharpen a complete draft instead of starting blank.
What changed.
- Proposal turnaround moved from days to hours
- Scope language became consistent across every author and practice
- Protective assumptions and exclusions stopped being optional
- Fewer change orders rooted in ambiguity — margin stays where it was earned
Construction · Estimating & Job Cost
The margin was set in a spreadsheet only one estimator understood.
The problem. At a construction firm, every bid started from a copy of the last workbook that looked close enough. Each estimator carried their own cost assumptions, quotes took days the schedule didn't have, and the number that won the job was never reconciled against what the job actually cost — so the same pricing mistakes were rebid for years.
The intervention. We built a reusable cost library from the firm's own history, put structured estimating on top of it — the pattern that became Estimator — with scenario comparison, margin visibility per line, and a clean path from estimate to proposal to contract.
What changed.
- Estimates assembled from a shared cost library instead of a personal workbook
- Bid turnaround dropped from days to same-day for standard work
- Margin became visible — and defensible — line by line before the bid went out
- Estimate-to-actual feedback made every job improve the next bid
- The repeated problem became a product — Estimator
Enterprise Data & Reporting
Three systems, three answers to the same executive question.
The problem. Enrollment, finance, and operational reporting each pulled from different systems with different definitions. Leadership meetings started with twenty minutes of arguing about whose number was right. An AI initiative was funded — and stalled immediately, because the data underneath had never been governed.
The intervention. A capability map exposed where the definitions diverged. We designed the integration strategy and governed data platform first, established shared definitions with named owners, and only then put reporting — and later AI — on top of it.
What changed.
- One source of truth with definitions leadership agreed to once, in writing
- Executive reporting that survives cross-examination
- Data foundation that made the AI initiative shippable instead of aspirational
- Integration patterns the internal team now applies without us