Decision models that hold up in the real world — before you commit patients, time, or capital.
Services & Pricing
Choose the depth of decision assurance you need—from Stage-0 (Logic Intake) to an audit-grade, reusable decision model (and optional managed runs).
We formalize decision logic and stress-test assumptions so decisions stay defensible under real-world variance.
Not a CRO. Not a biostats service. Decision logic, formalized.
Built for high-stakes decisions across research, development, operations, and governance.
We help teams decide what will actually hold up before they commit time, patients, or capital. If you want the same service described in your operating language, choose your context:
What you receive (inspectable artifacts)
- Assumptions & Fragility Register (MRR)
- Scenario Pack (Stress Tests)
- Decision Memo (review-ready)
Artifacts are inspectable (LaTeX/PDF), versioned, and reviewable—not slide decks or black-box outputs.
Engagement modes (how teams typically work with us)
Project-based
(most common)
Best when you have a specific protocol, pathway, study, or decision deadline. You pay per package (or fixed scope) and receive a deliverable you own—built around one clearly defined decision.
Subscription
(for ongoing teams)
Best when you run studies continuously and need repeated decision-focused modeling support across multiple questions, iterations, and datasets.
Not sure which fits? Start with Stage-0
Packages (pick the depth you need)
Most teams start with Stage-0, then select the package that matches the level of assurance and reuse they need.
Packages bundle multiple stages + integration/verification overhead; stage quotes apply only when scoping strictly by stage.
For one clear decision, fast de-risking, and a defensible logic package—without building a full reusable code asset. Designed to surface the highest-risk assumptions quickly—not to build a full reusable model.
Typical fit: decision memo + minimal stress tests (often maps to Stage-1, sometimes Stage-2) Includes: MRR (scoped) + scenario pack (lite) + decision memo Starting at: $1,900 Typical range: $1,900–$4,900
For a runnable model with deliverable code your team can execute (or we can run under NDA), so the decision logic can be re-run and inspected—not just described. Suitable for teams that need inspectable, rerunnable evidence—not just a static memo.
Typical fit: runnable code package + verification evidence (often maps to Stage-1–3) Includes: full MRR + scenario pack + deliverable code + verification evidence + runbook/configs Starting at: $12,500 Typical range: $12,500–$29,000
Small note: package pricing includes integration + verification overhead; stage-based totals may vary slightly.
For a long-term, auditable decision asset: deeper verification at the code level, a reproducibility pack, and a professional repo structure your team can reuse across studies.
Typical fit: Stage-1–3 + deeper verification/reproducibility Includes: audit-grade artifact set + deeper verification + reproducibility pack + reusable repo structure Starting at: $29,000 Typical range: $29,000–$75,000
Small note: bundle discount for Stage-1 → Stage-3 packages is available (typically 5–10%).
Quick chooser:
- Choose Lite if you need a fast decision memo (single decision + near-term deadline).
- Choose Standard if you need runnable, inspectable code (needs reruns / sensitivity + code ownership).
- Choose Advanced if you need an audit-grade reusable asset (multi-study reuse + audit/repro pack required)
For investment or governance decisions:
See the governance-grade audit pathway →
How we work (overview) — details live on the Process page
We follow a staged, lock-based workflow to prevent “assumption drift” between concept, math, and code
Stage-0
Logic intake + written verdict (entry gate)
Stage-1
Logic lock (assumptions + architecture)
Stage-2
Formalization (I/O contract + formulas)
Stage-3
Code & evidence (deliverable code + verification)
Stage-4
Managed runs (optional)
Stage-based pricing (reference)
Most teams use packages. Stage pricing below is mainly for technical scoping.
If you’re scoping by stages rather than packages, typical pricing ranges are:
- Stage-0: $0
- Stage-1: Starting at $1,900 • Typical $1,900–$4,900
- Stage-2: Starting at $3,400 • Typical $3,400–$8,900
- Stage-3: Starting at $6,900 • Typical $6,900–$19,000
- Stage-4 (Managed Runs):
- Option A (per run/batch): Starting at $950 • Typical $950–$4,500
- Option B (monthly managed execution): Starting at $2,400/mo • Typical $2,400–$9,000/mo
Managed runs include basic data sanity checks; full data cleaning is scoped separately.
Subscription (optional)
Support — $1,900/month
Best for: recurring questions, fast reviews, limited scenario work
For teams who model repeatedly across studies, iterations, and datasets.
Modeling Partner — $4,900/month
Best for: repeated use-case work + real iteration + light maintenance
Lab/Team Retainer — $9,500/month
Best for: multiple parallel threads, priority queue, deeper model/code continuity
What we need from you (inputs)
We don’t need identifiable data to start. Most decision-focused work begins with:
- protocol / study / pathway summary
- endpoint definitions + timing
- constraints (sites, recruitment reality, measurement limits; capacity/flow for pathways)
- a variable dictionary (even if rough)
Adding multiple use cases
(official rule)
Scenario Add-on
(same architecture)
If a new request adds scenarios, sensitivity checks, or reporting without changing architecture: $450–$1,900 (typical)
Scope Extension
(architecture changes)
If a new request changes the architecture (new state, new endpoint logic, PK/PD layer, new pathway structure, etc.), we reopen the lock:
- Stage 1 Re-lock: $900–$2,900
- Stage 2 Re-formalize: $1,400–$4,900
Use cases can be added either as Scenario Add-ons (same architecture) or as Scope Extensions (architecture changes). We confirm which one applies before quoting.
Data, IP, and privacy
- Stage-0 is anonymous and requires no NDA.
- From Stage-1 onward, NDAs and data-processing agreements (DPA) are available when appropriate.
- What we may publish (by default): sanitized, non-identifiable assets supporting transparency and reproducibility (generic model structure, reusable code components, synthetic examples, high-level descriptions).
- What we never publish (by default): your study data, identifiable information, confidential protocol details, internal documents, or anything revealing project identity or sensitive IP.
- Opt-out (client privacy mode): request at Stage-1 for no public release at all. This may affect scope, timelines, and pricing.
- Governance audits can be run in full client privacy mode by default upon request.
- Client confidentiality always overrides default transparency.
Payment & Delivery notes (short)
- Typical payment structure: 50% kickoff / 50% on delivery (per stage or package)
- Expedite option for tight deadlines: +25%
- Pricing is in USD (a USD/EUR toggle can be added)
Inspect our math
Public, versioned artifacts—reviewable by peers
Outputs are delivered in research-grade formats (LaTeX/PDF) suitable for citation and review—not marketing decks.
Not sure which stage fits your situation?
Most engagements begin with a quick Stage-0 review.
Stage-0 is the entry gate: a fast logic intake that routes you to the right package and the lowest-effort next step.
Stage-0 and audits test decision logic—not outcomes.
