LAIMDOTA / SUPERAI — Latvia's Civil Society AI Observatory: An Open Engineering Challenge

We stand on a civilizational boundary where the technological acceleration created by humanity has become a direct threat to its very existence. The world no longer belongs to decisions made by the human mind; it is drowning in a massive, uncontrollable deluge of data, where power and algorithms merge into a single, invisible mass. If civil society does not intercept control of the information field of its existence today, we will become passive witnesses to a self-enslavement managed by autonomous systems. This is not futuristic speculation—it is a question of survival for all of humanity and our nation. Technology must serve the consciousness of a free human, not become an instrument of power that surpasses and subjugates us.

Modern democratic society faces a volume of information that a human alone can no longer manage.

• Millions of procurement documents.
• State budgets.
• Scientific publications.
• Patents.
• Corporate legal links.
• Satellite observations.
• Environmental measurements.
• Regulatory amendments.

A human can manually check ten documents. AI can analyze millions. Therefore, we propose an open experiment: LAIMDOTA — Latvia's civil society AI observatory, which helps detect risks before they turn into crises.

It is not a "truth machine"

LAIMDOTA must not become a political censor or an automatic judge. Its task is not to determine who is good and who is bad, but to do what humans cannot do at scale: find anomalies and show a precise chain of evidence, so that humans can draw their own conclusions.

Key Architectural Modules

• Public Money Watch: Analyzes public procurements and budgets, looking for price anomalies, unusual markups, and hidden corporate links. The AI does not declare corruption; it points out: “There is a statistical anomaly here. Verify.”
• Technology Watch: Monitors patents, scientific publications, startups, and dual-use technologies to spot new social risks in a timely manner.
• Climate & Geoengineering Watch: Tracks publicly documented experiments, geoengineering research, and environmental measurements. The system distinguishes physically sound data from speculation, assigning an evidence level and the scope of the unknown to each claim.
• Regulatory Watch: Automatically monitors changes in laws and EU regulations, noticing rule adjustments before their consequences become obvious.
• Evidence Engine: For each significant conclusion, the system generates a transparent structure: claim -> sources -> primary evidence -> conflicting data -> level of confidence -> unknown variables.

Security and Operational Principle

The system's core logic is strictly limited:
DETECT → VERIFY → EXPLAIN → ALERT
It must never transition to the PUNISH phase. It does not perform political profiling and does not make autonomous decisions about people.

Engineering Framework (Open Stack)

The project is based on today's technology stack:

• Orchestration: LangGraph / LlamaIndex agent layer.
• Vector Database: Qdrant (for hybrid search of millions of documents).
• Graph Database: Neo4j (for Evidence Graphs structure).
• Data Pipelines: Apache Airflow / Temporal.io.

Open Question to Engineers

We do not claim that this architecture is ready. We submit it to a public engineering challenge.
Is it technically possible to connect finance, science, environment, and regulation into a single transparent data observatory?
If this idea is flawed—show exactly where the weaknesses of the architecture are. If it is possible—let's build a prototype.

Human determines values. AI searches for connections. Society retains control.