Lattice Graph × Builders Vision

Lattice Graph is a computational materials-discovery platform built around a governed knowledge graph that spans millions of compositions, mapping each one from atomic structure through experimentally measured and computed properties, synthesis routes, and patent coverage. When a company or investor brings us a materials thesis, we do not simply query a database — we run each candidate through multiple independent physics engines, including machine-learning interatomic potentials such as MACE and CHGNet alongside density functional theory, and we require consensus across those engines on phonon and thermodynamic stability before a candidate advances. That multi-engine validation step is the difference between a computational prediction and a defensible one. The second structural differentiator is what we call the negatives atlas: a curated set of more than 23,000 labeled failed experiments and kill edges that document what does not work and why. Nearly every public training corpus is built on published successes; failure results rarely clear peer review. Our atlas is predominantly internal, which means the pattern-recognition signal it provides — what chemical spaces are dead ends, which synthesis pathways consistently collapse, which property targets are physically unreachable at practical conditions — is not available to a model trained only on the open literature. For a diligence team evaluating an AI-materials startup's claimed data moat, that distinction is decisive. We also run a structured freedom-to-operate and patent-whitespace screen across more than 300,000 materials patents, integrated directly into the knowledge graph so that IP exposure surfaces alongside property evidence rather than as a separate, slower legal step. The result is a platform where a technical-diligence read — opportunity inventability, buyer landscape, computational stability, and IP position — can be produced in a bounded, traceable engagement rather than assembled piecemeal from a founder's deck and a literature search.

Builders Vision occupies an unusual position in the impact-capital landscape: it writes checks across catalytic and market-rate capital, which means its deal flow spans early-science bets and late-stage industrialization rounds in the same climate, food, and ocean thesis space. That breadth is a strength, but it creates a specific diligence vulnerability. AI-for-materials platforms and critical-minerals companies are technically dense, the fundraising cycle for these rounds has accelerated, and the decks that describe them have converged on a common vocabulary — foundation models, generative discovery pipelines, novel candidate screening — that sounds credible across the category regardless of underlying rigor. A generalist-leaning impact investor underwriting a materials-science claim without an independent technical read is essentially pricing risk it cannot see. The strategic fit with Lattice Graph is direct. We are not a consultant offering an expert opinion; we are a governed computational platform that can be pointed at a target company's claimed discovery space and return a traceable, scored answer. Is the claimed white space genuinely inventable, or already claimed? Does the company's model re-propose candidates that already sit on documented kill edges? Who are the funded buyers in this materials category, and does the target's commercial story hold up against observed procurement behavior? These are questions Builders Vision cannot answer from the founder's materials alone, and they are questions we are built to answer quickly and independently. The third dimension of fit is thematic alignment. Builders Vision's climate, food, and ocean programs intersect with precisely the materials categories where AI-discovery hype is highest and technical differentiation is hardest to verify from the outside: battery chemistries, critical-mineral separation and recovery, bio-based and ocean-derived materials, low-carbon cement and steel inputs. These are categories where Lattice Graph has deep knowledge-graph coverage and where our negatives atlas is most differentiated, because failure results in electrolyte stability and mineral-recovery chemistry are the data that the public corpus systematically underrepresents.

Builders Vision occupies an unusual position in the impact-capital landscape: it writes checks across catalytic and market-rate capital, which means its deal flow spans early-science bets and late-stage industrialization rounds in the same climate, food, and ocean thesis space. That breadth is a strength, but it creates a specific diligence vulnerability. AI-for-materials platforms and critical-minerals companies are technically dense, the fundraising cycle for these rounds has accelerated, and the decks that describe them have converged on a common vocabulary — foundation models, generative discovery pipelines, novel candidate screening — that sounds credible across the category regardless of underlying rigor. A generalist-leaning impact investor underwriting a materials-science claim without an independent technical read is essentially pricing risk it cannot see. The strategic fit with Lattice Graph is direct. We are not a consultant offering an expert opinion; we are a governed computational platform that can be pointed at a target company's claimed discovery space and return a traceable, scored answer. Is the claimed white space genuinely inventable, or already claimed? Does the company's model re-propose candidates that already sit on documented kill edges? Who are the funded buyers in this materials category, and does the target's commercial story hold up against observed procurement behavior? These are questions Builders Vision cannot answer from the founder's materials alone, and they are questions we are built to answer quickly and independently. The third dimension of fit is thematic alignment. Builders Vision's climate, food, and ocean programs intersect with precisely the materials categories where AI-discovery hype is highest and technical differentiation is hardest to verify from the outside: battery chemistries, critical-mineral separation and recovery, bio-based and ocean-derived materials, low-carbon cement and steel inputs. These are categories where Lattice Graph has deep knowledge-graph coverage and where our negatives atlas is most differentiated, because failure results in electrolyte stability and mineral-recovery chemistry are the data that the public corpus systematically underrepresents.

The most useful daily surfaces for a Builders Vision deal team are the knowledge-graph explorer and the opportunity and buyer intelligence views. The KG explorer lets an analyst walk the composition-structure-property-patent-recipe graph around whatever chemistry a target is building — inspecting the provenance chain behind a claimed property value, the cross-source agreement and disagreement flags across DFT calculations and experimental measurements, and the patent edges that reveal freedom-to-operate exposure. An opaque pitch deck becomes something the team can audit independently in an afternoon, with every claim traced back to its evidence source rather than to the founder's assurance. The opportunity index and funded-buyer leaderboard give the same team a ranked, screenshottable view of where the target's technology sits relative to the field, and who the realistic acquirers or offtake buyers are in that materials lane. For an active investor with a recurring pipeline of deep-tech deals, the highest-leverage workflow is batch screening: pushing a slate of prospective investments — or existing climate, food, and ocean holdings — through opportunity scoring, buyer affinity mapping, negatives and kill-edge checks, and patent whitespace in a single pass, producing a comparable technical scorecard across the portfolio rather than one-off reads assembled under deal pressure.

The natural engagement structure for Builders Vision is a scoped diligence read tied to a specific target company, delivered on a timeline that matches a typical deal process. A single engagement covers the full independent technical read: an opportunity-index position for the claimed discovery space, a funded-buyer affinity assessment, a negatives and kill-edge overfit test run against the target's announced candidates, a freedom-to-operate and patent-whitespace flag, and a provenance-grounded technical memo written for the investment committee rather than for a materials scientist. The lead hook for this engagement type is fast and bounded, with individual scoped reads estimated in the range of $30,000 to $60,000 — a framing figure, not a commitment, since final scope depends on the number of companies and the depth of the materials lanes in question. For an investor with ongoing deal flow in AI-materials and critical minerals, the engagement scales into a retainer-style diligence subscription that allocates a set number of scoped reads per quarter alongside standing access to the Opportunity and Buyer Intelligence engine and the Negatives and Eval-Data Atlas for the team's own first-pass screening. Portfolio monitoring — running existing holdings through updated kill-edge checks and patent-whitespace scans as the knowledge graph grows — can be added as a structured deliverable. All pricing figures here are estimates for planning purposes; final terms are set against Builders Vision's diligence cadence and the number of companies in scope.