Frequently Asked Questions
What Metal is, how the Context Graph works, and how it sits alongside the AI your team already uses
A context layer is a structured record of what your firm knows and what it concluded, sitting between your raw sources and the AI your team already uses. Generic AI is excellent on a single document and breaks the moment a question spans the whole firm, because conviction is built across every comparable and every risk you have seen before. Metal is that layer for private capital: every deal, company, and figure, plus the thesis, the risks flagged, and the decision you made.
No. Point generic AI at raw files and it re-derives structure on every question, which is why two versions of the same number can both come back cited. Metal resolves and deduplicates your sources into one Context Graph before anyone asks, so the same model works from one settled view of the firm. It’s not the model. It’s the data.
No. Metal’s open API and native MCP server bring your Context Graph into the tools your team has already chosen, so the AI you use keeps working and starts knowing your firm. It doesn’t replace the AI you use. It makes it know your firm.
Metal never trains on your data. The Context Graph enforces the permissions your firm already has in place, so a person asking through Claude or ChatGPT sees only what they could see in the source system, and every fact traces back to the document and page it came from.
You do not need to clean your data first. We start with a subset, a single live deal or one sector, structure it, and let your team judge the answers against work they already know. Messy, inconsistent sources are the normal starting point. The hard part of being AI-ready is ours: readiness is the product.
Join top firms redefining private capital with AI
Conviction, on every deal.
Metal Technologies Inc © 2026