We build companies that have earned formation through validation.
When Zero Vector starts, all that exists is potential. We identify high-potential people, identify problems internally, and match founders who demonstrate their capability to validated problems.
People. Problems. Evidence. Formation.
Companies shouldn't launch simply because someone has a cool idea or an impressive background. They should have to earn the right to launch through real evidence of founder capability and market need.
Identify and develop high-potential startup talent through real work.
Identify real problems that are extremely painful and impact a large number of people.
Validate founder capability through real work and problems through rigorous research.
Form companies only when the people, problem, and market have earned it through evidence.
Five stages to ensure company formation is driven by evidence and founder capability.
Problems and people are validated on parallel tracks. They converge only when a capable founder can be matched to a real problem.
Identify real, large, and significant problems that are worth deeper validation.
Identify people with the potential to build startup operator capabilities.
Kill problems that aren't real, extremely painful, large enough, or where a solution could realistically reach product-market fit in 18–24 months.
Validate the capability of potential founders based on how they perform with real startup work.
Founders are matched to problems based on their demonstrated capability and domain expertise. Problem validation continues in parallel.
Multiple solutions are identified. Solutions that lack a real wedge, lack defensibility, aren't technically feasible, don't deliver value to customers quickly, or aren't viable are killed.
Founders present their final case for company formation.
Identify real, large, and significant problems that are worth deeper validation.
Kill problems that aren't real, extremely painful, large enough, or where a solution could realistically reach product-market fit in 18–24 months.
Identify people with the potential to build startup operator capabilities.
Validate the capability of potential founders based on how they perform with real startup work.
Founders are matched to problems based on their demonstrated capability and domain expertise. Problem validation continues in parallel.
Multiple solutions are identified. Solutions that lack a real wedge, lack defensibility, aren't technically feasible, don't deliver value to customers quickly, or aren't viable are killed.
Founders present their final case for company formation.
Some of the core capabilities we focus on.
Capabilities are observed in real work. Confidence is gained by watching people demonstrate them across multiple cycles.
Converts feedback, evidence, and failure into upgraded judgment and changed behavior. Measured by the delta between prior action and next action — not by self-reported reflection.
Makes proportionate, evidence-aware decisions when information is incomplete. Separates fact from assumption, weighs evidence by quality, and acts without pretending to know more than they know.
Returns to clear thinking and useful action after setbacks, critique, or failed tests. Assessed by recovery behavior under real friction — not by motivational language or claims of grit.
Defines a specific, evidence-seeking problem before committing to a solution. Names who has the pain, what breaks, what it costs, and why now — and refines the frame as evidence accumulates.
Gathers truthful, behavioral evidence from users, buyers, approvers, and blockers without contaminating the signal. Probes for current workflows and costly action, not validation or compliments.
Designs a sequenced set of tests that attack the riskiest assumptions first, with evidence standards and decision thresholds defined before results arrive.
Identifies the few assumptions that must be true, ranks them by existential consequence, and pre-commits to the evidence that would force a continue, pivot, or kill decision.
Interprets accumulated evidence against pre-defined criteria to continue, change direction, or stop. Resists sunk cost, vanity traction, and goalpost-moving.
Distinguishes decision-useful evidence from anecdote, compliment, and vanity metric. Weighs evidence by source quality, behavior over opinion, and relevance to the decision at hand.
Uses pricing, budget discovery, and buying conversations as a learning system. Tests willingness to pay through real commitment and separates user desire from buyer urgency.
Control Layers for Complex Systems.
This is where we are currently focusing our problem discovery and validation work.
Control Layers for Complex Systems are software-led, infrastructure-light companies that help customers govern, secure, orchestrate, automate, verify, and make decisions across high-friction systems shaped by AI, regulation, cyber risk, infrastructure constraints, labor scarcity, climate volatility, and geopolitical fragmentation.
Define policies, permissions, thresholds, roles, and acceptable actions.
Protect systems, agents, workflows, data, and infrastructure from misuse or attack.
Coordinate work across systems, humans, AI agents, data sources, and physical assets.
Execute repeatable work under defined controls and human oversight.
Prove origin, status, compliance, action history, permissions, provenance, or outcome.
Help users make higher-quality decisions under uncertainty, scarcity, risk, or regulation.
Produce records that can survive audit, procurement review, litigation, compliance review, or board scrutiny.
Companies that earned formation.
Each company was matched to a validated problem and a capable founder.
Operations platform for emerging VC firms managing deals, LPs, and reporting.
Connecting early-stage founders with experienced product professionals for structured, incentive-aligned, product-focused advising.
The Product People Platform helps early-stage founders diagnose and improve the product decision-making loop through discovery, validation, UX, and problem framing before they overbuild, hire too early, chase the wrong market, or add fragmented product support.