All work

Case study

Live prototype

ValueDriver OS.

A greenfield operating system for PE value-creation teams — seeded with Red Iron Group's portfolio, built solo in days with Claude Code.

Build team
Solo
Concept to live
Days
Portfolio co's seeded
6
Primary surfaces
12

Context

Red Iron Group's thesis embeds operators — Builders — inside portfolio companies to drive value creation across technology, revenue, acquisitions, and operations. In practice that means a small team runs concurrent engagements across ~6 portfolio companies and tries to spot patterns across all of them.

The tooling for that job is off-the-shelf at best — general-purpose PM tools, spreadsheets, Notion pages. Nothing is shaped to the specific workflows of a PE operating team: cross-portfolio pattern recognition, value-creation roadmaps, executive cadence, a steady pipeline of ideas, clean handoffs between the fund and the company.

I wanted to answer a simple question: what would a purpose-built operating system for this team actually look like?

The premise

Before writing code I studied Red Iron's actual portfolio. ValueDriver OS is seeded with real company names, real sector shapes, and the four-pillar value-creation frame the fund operates inside. This isn't a generic PE product — it is a full-fidelity sketch of what a dedicated Red Iron operating platform could be.

The design premise: every surface is Claude-native. A Portfolio Assistant orchestrates an AI layer that does real operator work — finding opportunities, drafting requirements, generating reports, ingesting context from calls and emails. Learn Mode and guided walkthroughs sit alongside for onboarding. AI isn't a sidebar — it's the layer doing the work.

What I built

Twelve primary surfaces, solo, in days — built with Claude Code as the primary tool and deployed on Cloudflare Workers.

AI layer — built on topOpen DiscoveryCross-Portfolio PatternsReport GeneratorThesis HealthBaseline functionsDashboardCompaniesPipelineProjectsPatternsPlaybooksBenchmarksOperators
AI features sit on top of the baseline — every surface picks up Portfolio Assistant, Learn Mode, Walkthroughs, and pattern detection without the operator having to reach for them.
  • Portfolio Dashboard. Combined ARR, projected EBITDA lift, hold period, reporting status across all portcos.
  • Portfolio Companies. Per-company deep views — seeded with Red Iron's actual portfolio (Charter Impact, Graphite Financial, Finance Operations as a Service, Charter School Management, and more).
  • Cross-Portfolio Patterns. The pattern-recognition surface that defines strong operating teams — learn once, apply across the fund.
  • Deal Pipeline & Projects. Inbound diligence tracking and active value-creation workstreams alongside each portco.
  • Playbooks & Benchmarks. Reusable value-creation methods plus portfolio company performance against market benchmarks.
  • Operators, Knowledge Base, Academy, Integrations. Team roster and workload, shared reference material, onboarding curriculum, and the connective tissue to external systems.

Cutting across every surface: an AI layer that does real operator work — Open Discovery surfaces the highest-ROI opportunities, Cross-Portfolio Patterns finds reusable wins across portcos, Thesis Health tracks every pillar against live data, and Report Generator turns hours of prep into minutes. A persistent Portfolio Assistant ties it all together.

The AI layer in detail

Each feature targets a specific operator task. Each earns its place.

  • Portfolio Assistant

    What it does

    Conversational interface grounded in the full portfolio context — every portco, project, metric, and pattern alert accessible in one chat

    The edge

    Compound questions answered in seconds without a query or a reporting sprint

  • Open Discovery

    What it does

    Scans a portco's metrics, thesis health, and operational context to surface ranked value-creation opportunities

    The edge

    Operators walk into every CEO conversation already knowing the highest-ROI next move

  • Cross-Portfolio Patterns

    What it does

    Detects shared opportunities and risks across the fund — flags when the same automation or playbook applies to 2+ portcos

    The edge

    One operator's solution becomes reusable across the portfolio — wins compound at fund scale

  • Report Generator

    What it does

    Produces portco-specific reports on demand — current metrics, thesis health, project status, and recommended actions

    The edge

    Weekly operating updates and board prep collapse from hours to minutes

  • Thesis Health

    What it does

    Continuously assesses each investment-thesis pillar against live data and classifies as On Track / At Risk / Off Track

    The edge

    Early warning before any pillar drifts off thesis — no surprises in the LP conversation

Each AI feature in ValueDriver OS targets a specific operator task — turning work that used to take hours (or weeks, for onboarding) into something that takes seconds.

In the app

Click any screenshot to expand. Use ← → to navigate, Esc to close.

  • Portfolio Dashboard — combined ARR, value created YTD, projected EBITDA lift, and active projects across all portcos at a glance.

  • Per-company deep view — seeded with Red Iron's actual portfolio. Investment thesis, organizational pillars, key contacts, and the full operating picture on one surface.

  • A Claude-powered assistant lives persistently in the UI — ad-hoc analysis across the whole portfolio without leaving the page.

  • Toggle Learn Mode and hover any element — the model explains what you're looking at and why it matters to the operator.

What this demonstrates

ValueDriver OS is not a product I'm shipping to market. It's a proof — that I can research a specific fund's model, synthesize a dedicated product for their operating team, and ship a polished working prototype in days, solo, with AI as the build partner.

For a PE operating role, that is the loop in miniature — map a company, identify the leverage, ship the thing, show the CEO. The prototype exists so the same muscles I'd use inside a portfolio company are visible up front.