Deal work
40+ hours returned to a 12-person investment team every month
A firm-specific Carriage system took on new-deal review, data-room analysis, LBO modeling, memo drafting, and retrieval across the fund’s files.
- Live in week one, mapped to the firm’s strategy, templates, and formats
- Precedent models, memo structure, and screening logic encoded up front
- Extended with custom screening memos built on the fund’s proprietary data
- 15+ engineering hours dedicated every month to new builds
40+hours returned per monthDocumented investment-team capacity
Portfolio projects
$180M of annual construction spend put under invoice-level control
A PE-backed rural fiber operator was paying roughly 40 subcontractors against work orders and contract terms nobody could check at volume. Carriage ingests each invoice package, matches it to the work order, contract, and system status, then routes it to pay, review, or dispute.
- 15-week build · payback in 9 weeks, inside the first quarter after launch
- Invoice, timesheets, and backup structured and matched in ~12 minutes
- Billed labor, materials, scope, and rate card checked on every package
- Weekly leadership view of approved spend, flagged exceptions, and project status
$5.4Mannual run-rate EBITDA opportunityModeled portfolio-company impact
AI enablement
A secure AI foundation paired with hands-on adoption
A 90-day, firm-wide program that takes a team from ad hoc prompting to running and maintaining its own AI workflows — built on your existing Claude or ChatGPT enterprise accounts, or on accounts we procure and manage for you.
- Runs on your own Claude or ChatGPT enterprise plan — existing or procured for you
- Two live sessions a month — six across the program — on the team’s real work
- Monthly four-hour open office hours, plus email and phone support throughout
- 30+ purpose-built PE skills preloaded: screening, CIM review, diligence, IC memos, LBO
- Custom connectors to your drives and deal systems, on a managed org with access control
6working sessionsHands-on enablement