Calibration Sprint · First-in-Human Evidence Track

FIH Diligence Read

The first-in-human evidence track of the Calibration Sprint: a ten-business-day senior read of what a first-in-human oncology program has actually learned — and whether the next capital commitment is aimed at the uncertainty that matters most.

Every first-in-human program produces data. Far fewer produce decisions you can defend. The Read asks one question of your program, or of the asset you are evaluating: is the next planned experiment the most capital-efficient way to resolve the dominant value-driving uncertainty?

Book a 20-minute fit call Fixed fee, scoped after the fit call. A small number of engagements at a time.

What you walk away with

Gate ReadOne page. Six critical gates, published in full.
Scored Diligence ReadNine domains, scored and sourced.
Decision SynthesisBoard-ready. Fact, inference, uncertainty.
Readout callLive, adversarial.

Ten business days from kickoff. What each one is for →

Why This Exists

Early oncology data invites two opposite mistakes, and both are made at the same table.

Reading too much. A response rate at an uncontrolled dose, in a small selected population, becomes the thesis for the next round — and the doses, exposures and denominators underneath cannot carry the weight placed on them. Nothing about this looks like a failure while it is happening, which is what makes it structurally easy to miss.

Reading too little. A program with a coherent biological story and an interpretable early signal is discounted because its data package is disorganized — the evidence exists but was never assembled into a form a committee can trust. The asset is not weak; the presentation is.

Assets are mispriced in both directions by the same underlying failure: nobody separated what the data show from what the documents assert. And both mistakes are made at the same tables — term sheets, investment-committee memos, board approvals, partnering decisions. The Read is built for that table.

The Question the Read Retires

Given what this program has actually learned — at what confidence, and with what still unresolved — is the next committed dollar aimed at the uncertainty that most drives the asset's value?

That question sounds like an investor's question. It is also, precisely, the developer's question. That it reads naturally from both chairs is not a coincidence — it is the test of whether an FIH program was designed to learn or merely to advance. When the two chairs reach different answers from the same data, that disagreement is itself the finding: it usually means the program's story and its evidence have quietly diverged.

What the Read Delivers

Four artifacts, each built for a different reader and a different decision.

Every claim of consequence is retrieved, source-tiered and cited. Where the evidence does not support a recommendation, the Read says so — refusal is part of the output.

Gate Read one page

Six critical gates, each answered plainly: biological coherence, exposure plausibility, dose credibility, signal interpretability, safety maturity, and whether the remaining uncertainties are actually resolvable with feasible experiments. A single gate failure is named as such, with what it would take to reopen it.

Who reads it: whoever owns the decision — the partner, the CEO, the board chair. The whole picture on one page, before the detail.

Use it when: you need to know quickly whether there is a disqualifying problem, before spending committee time on the rest.

Wrong artifact if: the asset has not yet dosed a human and what you need is a design review rather than a read of results.

Scored Diligence Read nine domains

Nine weighted domains, from biological coherence to capital efficiency, each scored and sourced — with a confidence overlay that separates what the data show from what the documents assert. The score is a way of organizing the evidence and exposing where confidence is thin; it is not a prediction.

Who reads it: the diligence team, the clinical and regulatory reviewers, the analyst who has to defend the recommendation.

Use it when: the decision needs to survive someone else's scrutiny, and "we looked at it and liked it" will not.

Wrong artifact if: you want a number to put in a model. The scoring structures judgment; it does not replace it.

Decision Synthesis board-ready

What is known, what is inferred, what is unknown, and what could kill the asset — plus the specific next commitment the evidence does and does not support. Facts, inferences, uncertainties and recommendations are kept explicitly separate throughout, because most data rooms do not enforce that separation and most committee debates dissolve without it.

Who reads it: the board or investment committee — the people who were not in the working sessions.

Use it when: a tranche, a term sheet or a go/no-go gates the next step and the reasoning has to travel without you in the room.

Wrong artifact if: the audience needs the detailed technical review rather than the synthesis of it.

Readout Call live

A live senior pressure-test. Your thesis gets argued against, not summarized back to you.

Who joins: whoever actually owns the decision.

Use it when: always — it closes every Read. The written artifacts state the conclusion; this is where you attack it while it is still cheap to be wrong.

This is judgment work, not computation — but judgment that improves with structure: explicit gates a program must pass for the rest of the analysis to matter, and a stated hierarchy of which uncertainties drive value. The structure does not replace thirty years of oncology development decisions. It makes that read auditable, which is what a board, a committee or a skeptical co-investor is entitled to.

Who This Is For

A real capital or design decision on a first-in-human oncology asset, in the next 90 to 180 days.

Investors and family offices weighing a new commitment or a follow-on into an asset at or after first-in-human.
Biotech founders and CEOs preparing to defend an FIH design or emerging data to a board, an investor or a partner — or deciding what the next experiment should be.
Boards that want an independent senior read before approving the next tranche of spend.

Scoped typically to one asset and one decision, so an investor can commission it for each asset as it comes up.

You are a fit if the decision is real and imminent — invest or pass, follow on or step back, approve the next cohort or redesign it — and you want an independent senior read before the money moves. Where warranted, the Read is the natural on-ramp to a deeper engagement; a portion of the fee may be credited forward by agreement.

How It Works · 10 Business Days

Two weeks, on a fixed clock.

Day 0

Fit Call

Produces

Is the asset and the decision specific enough to read in two weeks? Conflict check, go / no-go, and scope.

Days 1–3

Read-In

Produces

Program materials and public-domain evidence; the working evidence surface.

Days 4–7

Gates and Domains

Produces

The Gate Read, then the scored domains with the confidence overlay.

Days 8–10

Synthesis

Produces

The Decision Synthesis and the readout call.

Delivered from Madrid (CET); the readout is scheduled in your time-zone overlap. Based in Madrid; much of this career was built in the Boston and Cambridge oncology cluster, at Millennium and Takeda Oncology.

Inputs Required

A specific asset and a specific decision. The Read is accepted only when both are sharp enough to resolve in two weeks.
Your data room or program materials, plus public-domain evidence — under a mutual NDA signed before anything confidential changes hands.
Never accepted: patient-level data or PHI, attorney–client privileged material, or pre-filing patent claims. If any of it arrives, work stops until it is withdrawn.

How Your Material Is Handled

Confidential work runs inside a protected channel under an explicit AI-governance charter, never in consumer AI tools. Where the analysis can run on de-identified or code-substituted material, it does.

The full operating standard →

Conflicts

OncAdios takes a small number of engagements in a narrow field. Before a Read is accepted, prior exposure to another program is checked for conflict. Where one exists the Read is declined and the reason given — that protects the company already engaged and the one asking.

What the Read Is Not

Not a regulatory submission, and not representation before any agency.
Not a validated predictive score or an algorithm. The instrument structures the review; the judgment applied through it is senior, human and accountable — thirty years of oncology development decisions, not a model output.
Not a guarantee of asset success or FDA acceptance — no honest reviewer can offer one.
Not clinical-trial operations, legal, valuation or financial advice.

How It Connects

One engagement, two tracks.

Which track?

One question decides it: what evidence has to carry your answer? If it rests on whether a proposed development or regulatory path is defensible, that is the development-path track. If it rests on what human data already collected actually support, that is the first-in-human evidence track, provided those data are mature enough to answer the question you are asking.

A Phase I program can sit in either. What decides it is not whether anyone has been dosed, but whether the human observations so far are mature enough, and central enough, to carry the answer. Where both matter, the fit call names the dominant track and says plainly what is being left out.

The two tracks differ in the object of judgment, not in geography. The development-path track reads a story against the expectations of a specific room — the agency or board it is about to enter, wherever that room sits. The first-in-human evidence track reads the evidence itself: whether a program has established a credible exposure range, whether its signal can be interpreted in the population it enrolled, whether its next experiment resolves the uncertainty that drives its value. Those questions are the same in Basel, Tokyo, Shanghai and Boston, because they are questions about the evidence rather than about a filing. The regulatory judgment applied through both spans four agencies: FDA, EMA, PMDA and NMPA, across seven approvals.

Programs are read wherever they are being developed. Delivered from Madrid, with the readout scheduled in your time-zone overlap.

Not sure which? The fit call sorts it. See the full ongoing roles and the operating standard behind every deliverable.

Start a Conversation

If the decision in front of you is real, the fit call is worth twenty minutes.

The fit call tests one thing: whether the asset and the decision are specific enough to read in two weeks. If they are, you get a scope and a fixed fee. If they are not, you will hear that too.

Book a fit call Ask a question first

Prefer email? jgn@oncadios.com

Deciding with colleagues? OncAdios at a Glance sets out the practice on two pages (PDF).