Skip to content

Built for private equity funds and their portfolio companies.

About Carriage

Private equity experience. Dedicated engineering. Systems that do the work.

Carriage was built for one market and one standard: complete complex private equity work to institutional expectations, then bring the same execution discipline to measurable value creation inside the portfolio.

Our operating principles

The output has to survive real review. The system has to survive real operations.

01

Complete the work—not just the first draft

The standard is a finished, editable, reviewable deliverable or a process that runs—not a clever answer inside a chat window.

02

Build around the firm—not a generic template

Every agent and operating system is mapped to the client’s data, precedent work, formulas, formats, controls, and review expectations.

03

Measure capacity and EBITDA—not novelty

Deal-team work is measured in capacity returned. Portfolio-company work is measured in cost removed, cycle time, quality, and EBITDA impact.

Why Carriage

Most AI vendors sell access. Carriage takes responsibility for execution.

The difference is not another model or interface. It is the ability to absorb complex work, encode the client's standards, return native deliverables, and keep improving the system after it goes live.

01

An email-native execution model

Send files and instructions through email, receive completed Office work product, and reply with comments to revise—without moving the team into another portal.

02

Full private equity workstreams

Carriage connects document review, extraction, research, modeling, memo drafting, and presentations so the output does not stop where the real work begins.

03

Autonomous systems tied to the P&L

Portfolio projects begin with a labor and process baseline, then replace the manual work with a controlled system whose savings can be made economically visible.

04

Dedicated engineering that stays involved

Carriage scopes, builds, integrates, maintains, and expands each client system as models, workflows, source systems, and team needs change.

Across the fund and portfolio

Three ways to start. One implementation partner.

The first engagement can begin with a live deal, an operating workflow, or a firmwide adoption need. The method remains focused, secure, and grounded in measurable work.

01

Custom AI Agents for Deal Work

Email the request and source files. Get back completed private equity analysis in your format.

02

Portfolio Company Projects

Replace manual operating processes with autonomous systems that remove cost and expand EBITDA.

03

AI Training & Enablement

Stand up Claude or ChatGPT properly across the firm, then train teams inside the work they actually do.

Trust by design

Built for sensitive work and accountable decisions.

Carriage handles CIMs, data rooms, and financial detail that cannot leak. Access controls, residency, retention, and traceability are designed into the workflow from the beginning.

NIST Cybersecurity Framework badge

NIST Framework

Security practices are aligned to the NIST Cybersecurity Framework.

Microsoft partner badge

Microsoft Partner

Carriage is a member of the Microsoft Partner Network.

Cloud professional certification badge

Cloud Certified

The team holds Google Cloud Professional certification.

CSA STAR Level 1 badge

STAR Level 1

Carriage is listed in the CSA STAR Level 1 self-assessment registry.

Zero-retention processing

Supported AI processing can operate under zero-retention terms without training on customer data.

Ownership

Your data remains yours

  • Full data portability
  • You set retention policies
  • Complete deletion on request
Control

Access follows your governance model

  • Role-based access control
  • Complete audit logging
  • SSO and MFA enforcement
Protection

Controls for sensitive financial work

  • AES-256 at rest, TLS 1.3 in transit
  • 100% US data residency
  • Runs on SOC 1/2/3-certified infrastructure

AI data protection

Your data never trains AI models

For supported AI processing, documents and financial data are handled ephemerally under zero-retention terms: not used for model training, not stored by model providers, and deleted after the request completes.

  • Zero Data Retention agreement in effect
  • Stateless, ephemeral processing architecture
  • No customer data stored by AI providers
  • Processing isolated within a Google Cloud VPC

View technical security detailsSecurity questions: security@carriagemodels.com

The best way to understand Carriage is to give us real work.

Start with a live deal workflow or a recurring portfolio process. We will scope the execution path, evidence standard, and measurable value.