You cannot fix the Hidden Year with a strategy memo. You fix it with infrastructure, sequence, and a self-funding mechanism that the CFO will defend in front of the board. This is the playbook.
Two installments ago, we established the diagnosis: the modeling problem at the heart of the modern PE operating model is structural, not cyclical. Last week, we itemized the bill: every portfolio company in the fund is paying a two-to-six-million-dollar Hidden Year Invoice that no one has been asked to add up.
This week, the prescription. Not at the strategy layer — there are enough strategy decks. At the operational layer, where the engineering decisions become financial decisions and the financial decisions become the multiple.
The Universal Valuation Formula
Every boardroom conversation about AI investment ultimately resolves to the same equation:
Enterprise Value = Revenue × EBITDA Margin × Valuation Multiple
The formula is not complex. What is complex — and what most investment memos miss — is that an AI program affects all three variables simultaneously, on different timelines, with different evidence requirements. The underwriting that makes all three impacts visible is the one that earns the re-rating rather than hopes for it.
Lever 1 — Revenue. Green Belt territory. Churn prediction, personalization, intelligent pricing. At Amazon Rentals, we took the textbook rentals business from $35M run rate to over $300M — laying the foundation for what became a $1B+ commerce capability. That was not a projection. That was AI-driven personalization and intelligent inventory in production, generating verifiable revenue impact, on a timeline measured in quarters not years.
Lever 2 — EBITDA Margin. White Belt territory. At Visible (Verizon), we drove a 75% reduction in cost-to-serve through Human-Powered AI and AI-Powered Humans — the architecture covered by US Patent 12,056,644 B2, deployed in 2020, two years before GenAI made these concepts mainstream. Every dollar of OPEX saved drops to EBITDA. At your current valuation multiple, that OPEX reduction has already created enterprise value the IC can read off the page.
Lever 3 — Valuation Multiple. Black Belt territory. Traditional enterprises trade at 6–8x EBITDA. Technology-enabled enterprises trade at 8–14x. AI-native enterprises trade at 14–40x. The MuShuHaRi journey is the path from the first category to the third. The delta, multiplied by the EBITDA, is the enterprise value number that turns "should we approve this AI investment?" into "how do we accelerate this timeline?"
AI investment is not a cost to be minimized. It is a multiple to be expanded. The investor who internalizes that distinction underwrites a different deal than the one who is still asking why the technology budget is so high.
The Self-Funding Envelope
In conventional AI deployment, every stage requires a fresh capital ask. The CTO walks into the board, asks for $3M to build infrastructure, gets pushed back, and eventually gets a fraction of what was needed. The infrastructure underdelivers. The next ask is harder. By Year Three, the board's appetite is exhausted and the AI program has shipped two pilots.
The Self-Funding Envelope inverts this. The White Belt stage — cost liberation — is funded against the verified, CFO-confirmed cost-to-serve reductions it produces in the back office. Those savings, once verified, fund the Green Belt stage — commercial AI — in full. The verified revenue uplifts from the Green Belt fund the Black Belt stage. The program funds itself, stage by stage, against measured outcomes rather than projected ones.
The CFO is not being asked to take a leap of faith. The CFO is being asked to validate a savings number that has already happened, and to release the next tranche of capital from those verified savings. This is the conversation the CFO has been waiting their entire career to have with the AI program. It is also the conversation the IC has been waiting to read in a memo.
The DouJou compression
Under the conventional Build path, the infrastructure required to move a portfolio company from Mu to Ri takes twelve to sixteen months and runs $1.7M to $3.2M in senior engineering time on work that generates zero competitive differentiation. None of that infrastructure is proprietary. Every portfolio company in your fund builds the same plumbing. Industrialized waste, in slightly different colors.
DouJou deploys the complete MuShuHaRi infrastructure stack — White Belt through Black Belt — inside the portfolio company's own AWS VPC (Azure in Q3 2026, GCP in Q4 2026), with their data never leaving their environment. Six to nine weeks instead of twelve to sixteen months. Generalist developers instead of PhD-level specialists. The model layer is agnostic by design — GPT, Claude, Gemini, Llama, whatever the lab releases next quarter — because the model is now the replaceable component, not the lock-in. The data, the pipelines, the governance, the proprietary intelligence: those stay with the portfolio company.
Build vs. Buy vs. Third Way
| Dimension | Conventional Build | SaaS (Buy) | DouJou |
|---|---|---|---|
| Speed | 12–16 months | Days | 6–9 weeks |
| Data sovereignty | Maximum (your infra) | Minimum (vendor cloud) | Maximum (your VPC) |
| Talent required | PhD-level specialists | Business users | Generalist developers |
| Competitive moat | Strong (proprietary build) | None (shared platform) | Strong (your data, shared infra) |
| Maintenance burden | Permanent | Vendor's | Automated |
| Model flexibility | Whatever you build | Vendor's model | Model agnostic |
The compression is not marginal. It is the difference between the multiple expansion happening on your watch and the multiple expansion happening on the buyer's watch.
The fund-level move
If you are an LP, an IC member, or an operating partner, here is the operational sequence I would recommend running across your existing book — not as a thesis exercise, but as a portfolio management exercise:
Step 1 — Run the Mirror Test on every position. The MuShuHaRi diagnostic is a 30-question instrument that produces a 1–10 honesty-weighted readiness score per portfolio company. Your operating partners have been telling you their companies are at Green Belt. The diagnostic will tell you which ones are. Expect the gap to be substantial. That is the data.
Step 2 — Calculate the Hidden Year Invoice per company. Four lines, validated by the portfolio company CFO. Failed Pilot Tax + Infrastructure Opportunity Cost + Production Delay Cost + Talent Attrition Cost. Sum across the portfolio. That number is the floor of the operational waste your operating partners are currently managing past.
Step 3 — Identify the highest-leverage three positions. Not the largest positions. The positions where the gap between the underwritten exit multiple and the current MuShuHaRi stage is widest. These are the positions where re-rating capital generates the highest IRR contribution per dollar deployed.
Step 4 — Deploy the Self-Funding Envelope inside the three lead positions. White Belt first. Cost liberation. CFO-verified savings. Use the savings to fund Green Belt. Verified revenue uplifts. Use those to fund Black Belt. The capital plan does not require fresh equity at any stage.
Step 5 — Standardize the infrastructure layer across the fund. Deploying the same infrastructure stack across thirty portfolio companies, instead of having thirty engineering teams each rebuild it, eliminates the duplicated Hidden Year. The aggregate savings across the portfolio in Year One alone is in the tens of millions, before counting the multiple expansion the savings finance.
A fund that runs this sequence on its existing book in 2026 is a fund that re-underwrites its own positions at 2028 multiples. Without writing a new check.
What this looks like to the LPAC
In the old narrative, the GP arrives at the annual meeting with a portfolio update that defends marks against a market that has compressed software multiples by 30%. The LPAC asks pointed questions about exposure to AI displacement. The GP describes operating partner workshops and AI strategy reviews. The LPAC nods politely and asks the same questions twelve months later.
In the new narrative, the GP arrives with a portfolio-level MuShuHaRi diagnostic, a portfolio-level Hidden Year Invoice, a portfolio-level capital plan that does not require additional equity, and a per-company timeline showing which positions cross each threshold — Mu-to-Shu, Shu-to-Ha, Ha-to-Ri — and what multiple impact each crossing carries. Each threshold is gated by evidence the CFO has confirmed. That is not a strategy update. That is a re-underwriting of the entire fund, presented as a sequence of evidence-gated decisions rather than a sequence of capital asks.
The honest closing
Two final notes, in the spirit of practitioner candor. First, the framework is not magic. MuShuHaRi catches the failure modes most enterprises walk into unconsciously. It does not substitute for management quality, market position, or the basic operational rigor that distinguishes a good portfolio company from a poor one. A portco with weak management at Mu will not become a strong company at Ri. It will become a more expensive Mu.
Second, DouJou is the infrastructure layer, not the strategy. The framework tells the company which capabilities to build in which order. DouJou delivers the infrastructure on which those capabilities run. The proprietary intelligence — the domain knowledge, the customer understanding, the competitive insight — is what the portfolio company brings, and is what gets re-rated by the market.
The funds that AI-proof their portfolios in 2026 will write the IRRs that define the asset class for the next decade. The funds that don't will spend the same decade explaining why "AI was harder than we thought." Both narratives are already being drafted. The choice is which one your fund tells.
I will leave you with the question every IC should be asking on every position, starting Monday morning: What stage of MuShuHaRi is this company actually at — not the deck answer, the diagnostic answer? What is the company's Hidden Year Invoice this quarter? And how does the next twelve months of capital deployment compress the journey from where they are to where the multiple expansion lives?
Three questions. They will reset every conversation in the operating partnership. That reset is the work.
The Ikigai of capital is purposeful capital. Capital that builds the Dojo rather than runs the next pilot. Capital that funds the foundation rather than the press release. Capital that earns the multiple rather than hopes for it.
The Dojo is open.
◆ ◆ ◆
— Himanshu Niranjani
Founder, BeHuman Capital · Architect, DouJou · doujou.ai