Programs the standard model couldn't fund.

Pillar I

Platform & venture architecture.

Building the structure that lets programs exist which could never carry their own cost alone.

Cost architecture

Case 01

The Constellation Model

Platform architecture for ultra-rare disease · originated and owned as strategic design methodology

Challenge: Ultra-rare genetic diseases fail a simple economic test: each affects too few patients to carry the fixed cost of its own program. Under the one-disease/one-asset model the science is ready long before the financing ever will be. The blocker was never biology, it was cost structure.
Approach: Show how it was designed
A platform-plus-program architecture in which a single delivery-and-editing core is built once, validated once, then carries each new indication as a marginal addition rather than a full build, the “new star” joining an existing constellation. Around it: the separation of design from independent verification, a per-locus method for the program-specific layer, and the amortization logic that spreads shared preclinical, manufacturing and regulatory work across the cluster instead of duplicating it.
Outcome: A coherent architecture in which multiple hepatic metabolic indications run off one shared core, with cost per additional program falling as the constellation grows. It became the design layer beneath a live rare-disease venture, and separately a transferable methodology now applied for clients facing the same fixed-cost trap elsewhere.

Platform strategyRare diseaseGene editingCost amortization

Entity & governance map

Case 02

Dual-entity structuring with reversionary IP

Nonprofit / for-profit venture design and board governance

Challenge: A rare-disease platform needs two things that pull in opposite directions: philanthropic and public capital, which require mission lock and charitable status; and commercial capital and speed, which require an equity vehicle. Choosing one forecloses the other. Worse, founders who build the venture before it exists create related-party exposure the moment they are compensated for that work.
Approach: Show how it was designed
A structured nonprofit-versus-for-profit analysis, resolving to a paired structure: a charitable foundation holding mission control, a public benefit corporation holding execution and capital, bound by reversionary IP covenants so assets return to mission if commercial stewardship fails. Then the governance to survive scrutiny: disinterested directors approving all related-party terms, founder recusal, nothing contingent on a pending award, and a chain of title separating the strategic design layer from the applied execution layer.
Outcome: A structure that can accept philanthropic, public and private capital without one channel poisoning another, with a defensible founder-compensation framework and a provenance record clean enough to withstand diligence. The ownership-tiering method it produced is now a standalone client instrument.

Entity designBoard governanceIP provenanceRelated-party §4958

Capital stack · illustrative

Case 03

Funding-stack & applicant-routing architecture

Non-dilutive capital strategy across federal, state, international and philanthropic channels

Challenge: Non-dilutive funding is not one market but a dozen, each with its own rules on who may apply, where the work must be performed, and what entity type qualifies. Most ventures pick one channel, size the raise to it, and leave the rest untouched, because they cannot answer a deceptively simple question: which of our entities should be the applicant?
Approach: Show how it was designed
An applicant-entity routing framework mapping each channel against entity type, in-state nexus, academic-partnership requirements and cost-allocation rules, then sequencing them so no award compromises eligibility for the next. Where a channel required an academic prime, the pairing was structured; where it required in-state presence, the applicant was routed accordingly. Formation capital was treated as its own problem, the stage before the venture can apply for anything.
Outcome: A layered stack roughly double the ceiling of the single-channel default, with each layer’s eligibility protected rather than traded away. The routing logic transfers directly: it is the analysis any platform venture needs before it writes its first application.

Non-dilutive capitalSBIR / STTRGrant strategyEntity routing

Pillar II

Regulatory, evidence & development strategy.

Routes to approval for populations conventional trial infrastructure was never built to serve.

Regulatory sequencing

Case 04

Cluster IND & the sponsor ladder

Regulatory strategy for n-of-few populations

Challenge: A child with an ultra-rare variant may need treatment in months, not years. The fastest route, a single-patient expanded-access filing, treats each patient as a one-off and builds nothing toward an approvable product. The route that builds toward a label cannot move at the speed of the disease. Teams are forced to choose, and usually choose wrong, because the choice is framed as binary.
Approach: Show how it was designed
Reframe it as a ladder, not a fork. Each rung is a distinct sponsor, investigator, academic bridge, foundation, then commercial, holding the filing where it can act fastest, with each filing built so its data and chemistry carry upward to the next rung. The shared core makes this possible: it is the same platform underneath, so a cluster or master-protocol filing can amortize the common module across indications while each variant enters through its own gate.
Outcome: A route in which urgent single-patient treatment and a genuine path to approval stop competing, the first patient’s filing becomes the foundation of the cluster rather than a detour from it. Reinforced by comment submissions into live federal regulatory and oversight proceedings.

IND strategyExpanded accessMaster protocolRegulator engagement

Evidence architecture · illustrative

Case 05

Real-world evidence & long-term follow-up

Evidence architecture for small-n, permanent-effect therapies

Challenge: A one-time genetic edit is permanent, so regulators require years of follow-up. But the treated cohort may be a handful of patients scattered across continents, and the natural-history comparator often does not exist. Conventional trial infrastructure is built for populations these diseases will never have. Once the trial closes, the data thins exactly when the long-term safety question becomes answerable.
Approach: Show how it was designed
A connected evidence layer treating the registry, the natural-history dataset and the mandated long-term follow-up as one continuous system rather than three disconnected obligations. Patients are enrolled into the evidence architecture before they are treated, so each contributes both as their own control and as a durable longitudinal record. Data standards were set to be portable across sites and jurisdictions from day one, retrofitting them later is the failure mode designed out.
Outcome: An evidence system in which the small cohort becomes an asset rather than a limitation. Regulatory follow-up, payer evidence and natural-history control are satisfied by one architecture instead of three, and the data survives the closure of any individual trial.

RWENatural historyRegistry designLTFU

Decision framework

Case 06

Incentive & classification strategy

Regulatory classification, priority review vouchers, and testing the load-bearing assumption early

Challenge: Rare-disease business cases routinely rest on an incentive the model treats as a certainty, most often a priority review voucher booked at a confident nine-figure value. Programs get financed on that assumption. When the voucher does not arrive, or arrives worth a fraction of the model, the venture discovers its load-bearing assumption only after it has already built on it.
Approach: Show how it was designed
TMDG’s “what-would-have-to-be-true” discipline: rather than forecasting the incentive, invert the question, what conditions must hold for the strategy to survive without it? That meant analysing how the therapy is classified in the first place, because classification determines which incentives are even reachable, then stress-testing the base case at zero voucher value and modeling the alternatives honestly. The same method drove submissions into live federal proceedings, where the classification rules themselves are still being written.
Outcome: A strategy that holds at a zero-dollar voucher, treating any incentive as upside rather than foundation. Alongside it, a repeatable framework for finding the single assumption a business case cannot survive losing, before the capital is committed.

PRV strategyClassificationWWHTBTPolicy comment

Pillar III

Enterprise leadership & team building.

Building the function, and the people in it, while the programs are already running.

Operating model

Case 07

Standing up a portfolio & program function

ASX-listed global vaccines manufacturer · Head of Portfolio & Program Management; VP, mRNA Program

Challenge: A global vaccines organization moving into mRNA had strong functions and no connective tissue between them. Decisions were being made in five places at once, program-level accountability was ambiguous, and no single view of the portfolio existed to tell leadership what to fund, slow, or stop. The function had to be built while the programs were already running, there was no pause.
Approach: Show how it was designed
Established the portfolio and program management function from the ground up: hired and developed the team, defined decision rights explicitly rather than leaving them to seniority, and built the governance cadence that gave leadership one portfolio view instead of five functional ones. Concurrently led the mRNA program itself, so the operating model had to survive contact with a real, fast-moving, technically hard program, not just a slide.
Outcome: A functioning portfolio office with a team built and retained, unambiguous accountability at program level, and a single governance forum where trade-offs were actually made. It was designed to keep working after any one program moved on, which is the real test of whether an operating model was designed or merely imposed.

Org designTeam buildingPortfolio governancemRNA

Portfolio view · illustrative

Case 08

Corporate strategy at platform scale

Nasdaq-listed mRNA therapeutics company · Chief Strategy Officer

Challenge: A clinical-stage platform company with more credible directions than capital to pursue them. Every program had a champion and a defensible case; the constraint was not idea quality but the discipline to concentrate. Public-market scrutiny compressed the timeline for getting the answer right, and platform companies fail more often from diffusion than from any single program failing.
Approach: Show how it was designed
Built the strategy function around explicit trade-off logic rather than advocacy, mapping programs against strategic fit and value, forcing the question of what the company would have to believe for each to earn its funding. Partnership and business development were treated as a portfolio instrument, not a fundraising afterthought: the right partner validates a direction and de-risks a channel, not just supplies capital. That analysis fed the board and the investment narrative directly.
Outcome: Concentration where it counted, and honest early kills where it did not. The strategy narrative held up under investor and board scrutiny because the trade-offs behind it were visible rather than assumed.

Corporate strategyBusiness developmentBoard & investorsa-mRNA

Capital architecture · counterparties withheld

Case 09

Cross-border partnership & capital architecture

Public–private infrastructure partnership · approximately US$1.3B · outside life sciences

Challenge: A public–private infrastructure partnership spanning two jurisdictions. The parties’ incentives, legal regimes and political timelines did not naturally align. Public and private capital each required assurances the other could not straightforwardly give.
Approach: Show how it was designed
Architected the partnership structure: the governance bridge between the parties, the alignment of public and private capital into a single financeable form, and the sequencing of commitments so no party had to move irrevocably before the others were bound. Counterparties, financing mechanics and execution status remain confidential.
Outcome: A structured partnership at approximately US$1.3B in infrastructure, counted separately from the $7B of program and portfolio value above. It is included because the structuring capability transfers across sectors. Aligning mission capital, public capital and private capital under one governance frame is the same problem whether the asset is infrastructure or a rare-disease platform.

Public–private partnershipCapital structuringCross-borderGovernance

Pillar IV

The instruments.

Advice expires; instruments don’t. Every engagement leaves behind a working tool the client owns and keeps using. Each was built from that client’s answers, which is why the work starts with a conversation.

Weighted decision engines

0–100 composite scoring with adjustable weights, tier bands and color thresholds, for opportunity triage, partner selection and portfolio prioritization. The weights are exposed so the client can argue with them.

IP provenance & chain-of-title register

A tiered record of what was originated, what was borrowed and what was jointly built, with the chain of title traced for each. The instrument diligence asks for and most ventures cannot produce.

Compensation & consideration models

Multi-vehicle stacks, reimbursement, milestone, retainer, deferred-to-equity, founder equity, royalty, modelled as a waterfall, with a matrix showing what each vehicle costs the venture and pays the principal.

Program plans & decision-rights maps

Integrated development plans with explicit gates, paired with RACI maps that say who decides, not who attends. The pairing is the point: a timeline without decision rights is a wish.

Pipeline & performance dashboards

Auto-recalculating trackers built around a stated north star, qualified pipeline and closed work, not vanity volume. Refreshed from native platform exports, so they survive without a data engineer.

White papers & convening architecture

Published positions that originate work rather than decorate it, each paper mapped to the audience it must reach and the conversation it must start, with convenings designed as the follow-through.

Structures built where the standard model fell short.