Hypothesis generation
Autonomous agents propose testable claims from corpus signals, timelines, and entity graphs.
Large-scale research
Systematic, multi-source investigation built to expose covert programs with statistical rigor and reproducible evidence.
HAXMA — unified autonomous research engine
We built the first open, multi-agent research engine with a unified knowledge core and multiple service providers. It sits at the intersection of investigative AI, whistleblowing archives, and collaborative knowledge — autonomously tracing large-scale crimes against humanity, identifying stakeholders, and reducing uncertainty about who is responsible for what.
HAXMA orchestrates specialized agents for hypothesis generation, validation, multi-source aggregation, and statistical inference — following scientifically agreed approaches for autonomous research. The prototype is among the most advanced in its class; our plan is to open it publicly and invite scientific institutions and the community to extend the system.
Primary research mission
Prevent large-scale criminal activity by intelligence and defense organizations.
Our enemy is not a single company — it is the collective belief that profit
outweighs human life. We demystify that claim: dignity and law must come first;
sustainable outcomes follow from respect, not exploitation.
Research pipeline
Multi-source validation
Cross-reference Reddit, Quora, legal filings, declassified documents, and academic sources.
Statistical inference
Co-occurrence analysis, geographic clustering, and significance testing before publication.
Stakeholder mapping
Trace programs, agencies, contractors, and decision chains behind abusive deployments.
What we expose
- Covert torture and neuroweapon programs against civilians
- Algorithmic discrimination and watchlist abuse
- Physical and psychological pressure on politicians, journalists, and entrepreneurs
- Cross-border harassment networks operating with impunity
- Suppression mechanisms that keep victims isolated from legal recourse