CONTACT
CONTACT
CONTACT
NEURAL · Institutional intelligence

Institutional knowledge. Investment conviction.

NEURAL is built to turn institutional knowledge into investment intelligence. Source-grounded evidence, specialist methodologies, editable financial models and independent challenge come together in a complete underwriting process—with material decisions remaining in human hands.

Explore the flow
Explore the flow
The eight products
The eight products

Evidence traceability

Human investment authority

Explore the connected layers
The platform

Four connected layers. One institutional purpose.

01 · Knowledge

A Library that retains institutional knowledge.

The source-grounded Library organizes reusable knowledge from authorized sources and the user’s experience across sectors, companies, transactions, financing structures, instruments, valuation and underwriting methods. The Librarian identifies and delivers applicable knowledge, support, qualifications and missing evidence.

LibraryLibrarianSource support
02 · Investor products

Outcomes an investor can commission.

Investment screens; institutional underwriting and full IC delivery; business, equity and claim valuation; financing and capital-stack design; instrument briefs; downside and recovery analysis; independent review; and affected revisions or what-if analysis.

Eight outcomesInvestor mandateComplete delivery
03 · Execution

The appropriate route for each assignment.

RAPID-lite is the lean route for screening and focused work. Modular makes specialist responsibilities explicit. Canonical is the integrated reference workflow with durable state, controlled revisions and auditable records. The work determines the route; all three do not run automatically.

RAPID-liteModularCanonical
04 · Access

The interface is replaceable.

Plugin and API / MCP access connect users to the same underlying intelligence. A desktop application is being developed. Across interfaces, the Library, analytical process, models, review and product standards remain underneath.

PluginAPI / MCPDesktop in development
Beyond the first draft

A model-backed process. A reviewed investment product.

An investment case should reveal how the business earns, how cash reaches the investor, where protection can fail and what would change the recommendation. Each specialist contributes to that decision, not simply to another draft.

Editable financial models

Connect financial spreading to a model that can be inspected, revised and used to test the investment case.

Model-backed analysis

Valuation and financing

Bring business, equity and claim valuation together with financing structure and capital-stack analysis.

Methods made explicit

Separate economic review

A distinct analytical review checks the economic meaning of the model, its assumptions and its limitations.

Independent economic view

Qualitative challenge

An independent challenger interrogates the investment case, contradictions and unsupported assumptions.

Constructive disagreement

Controlled revisions

Preserve analytical history and evidence custody while revising the parts affected by new information or a changed assumption.

Versioned corrections

Complete IC delivery

Coordinate the memo, model, supporting analysis and presentation into an investment-committee package.

One coherent product

Reusable intelligence

Focused tools automate valuable underwriting tasks. NEURAL is designed to turn reusable institutional intelligence into complete investment products through a model-backed, independently reviewed and traceable process. Those tools can work alongside it, or become focused actions within the broader platform.

Purpose and progress

What NEURAL is building toward.

No. The prototype has demonstrated much of the end-to-end analytical and product chain. Hosted access and repeatable ordinary execution are still being completed. It is not yet fully rolled out across the team.

The Library retains reusable, source-grounded knowledge and authorized experience. The Librarian helps deliver the right methods, supporting sources, qualifications and missing evidence to the relevant participants in each underwriting.

The product direction includes live interaction with the AI underwriting team: challenge assumptions, answer questions and discuss technical matters in writing. Voice interaction is part of the intended experience; availability depends on the implementation and access channel.

Yes. Financial spreading, first-pass modelling, memo drafting and financing-grid tools can operate alongside NEURAL or be incorporated as focused actions. The distinction is the broader intelligence and coordination layer, not a requirement to replace every useful tool.

The underlying architecture is intended to support other complex, high-judgment work by changing the knowledge base, methods, participants, controls and required products. Institutional credit is the application described here; the same design principles remain: traceability, explicit methods, review and human authority.

No. The prototype has demonstrated much of the end-to-end analytical and product chain. Hosted access and repeatable ordinary execution are still being completed. It is not yet fully rolled out across the team.

The Library retains reusable, source-grounded knowledge and authorized experience. The Librarian helps deliver the right methods, supporting sources, qualifications and missing evidence to the relevant participants in each underwriting.

The product direction includes live interaction with the AI underwriting team: challenge assumptions, answer questions and discuss technical matters in writing. Voice interaction is part of the intended experience; availability depends on the implementation and access channel.

Yes. Financial spreading, first-pass modelling, memo drafting and financing-grid tools can operate alongside NEURAL or be incorporated as focused actions. The distinction is the broader intelligence and coordination layer, not a requirement to replace every useful tool.

The underlying architecture is intended to support other complex, high-judgment work by changing the knowledge base, methods, participants, controls and required products. Institutional credit is the application described here; the same design principles remain: traceability, explicit methods, review and human authority.

No. The prototype has demonstrated much of the end-to-end analytical and product chain. Hosted access and repeatable ordinary execution are still being completed. It is not yet fully rolled out across the team.

The Library retains reusable, source-grounded knowledge and authorized experience. The Librarian helps deliver the right methods, supporting sources, qualifications and missing evidence to the relevant participants in each underwriting.

The product direction includes live interaction with the AI underwriting team: challenge assumptions, answer questions and discuss technical matters in writing. Voice interaction is part of the intended experience; availability depends on the implementation and access channel.

Yes. Financial spreading, first-pass modelling, memo drafting and financing-grid tools can operate alongside NEURAL or be incorporated as focused actions. The distinction is the broader intelligence and coordination layer, not a requirement to replace every useful tool.

The underlying architecture is intended to support other complex, high-judgment work by changing the knowledge base, methods, participants, controls and required products. Institutional credit is the application described here; the same design principles remain: traceability, explicit methods, review and human authority.

Built for investment judgment

Intelligence that compounds.

NEURAL remains in prototype. The goal is a repeatable institutional process with meaningful specialist contributions, evidence custody and preserved human authority over material conclusions.

Explore the flow
Explore the flow
Back to the lab
Back to the lab

Create a free website with Framer, the website builder loved by startups, designers and agencies.