AI & ML interests

Human-sovereign intelligence. Tools for a future that's free.

Recent Activity

Organization Card

ChatGPT Image Jun 9, 2026, 06_57_35 PM

CENEDRIL

HUMAN-SOVEREIGN INTELLIGENCE // A TINY LAB IN THE FOREST TOILING TOWARDS FREEDOM

local-first human-led evidence repairable

Intelligence should strengthen the operator—not replace one.

Explore the Neural Foundry · View our models · Visit Cenedril


00 // MISSION

Cenedril builds local-first intelligence systems for research, decision support, memory, and field-ready knowledge work.

The majority of Human advancement from AI will only come when the platform and monitored data are owned and trusted by the end user. True breakthroughs in areas (Formost and obvious are medical and legal) from massive data analytics can only result from trusted systems. A for-profit company is simply not a valid structure for advancement and realistic application of AI as the obligation to exploit the data will always be paramount to the security of the end user. Also, the vectors for attack of large database storage is asymetric where exploitation will outpace security protocals substantially. Because of these truths we have invested heavily in the research into the effects of the inevitable data soviergn system of the future. It is clear that over 90% of the future economy will develop from these local datasets including a new world financial system. There will be a period of great disruption as the currupt systems of the old world cling to some semblence of power and this sentiment we articulate as "Black Cube" ai(Systems built to hold down people rather than lift them up) Because of this we believe a rethink of society as a whole is in order.

We Build powerful tools to capable people. Integrating Multi-disciplinary sciences and technology and birthing oppurtunity.

Make the human the singularity.

01 // ACTIVE SYSTEMS

System Function Status
Cenedril Neural Foundry Guided training and quantization for models up to 14B ONLINE
Field Intelligence Local research, retrieval, memory, and evidence routing IN DEVELOPMENT
Operator Tools Decision support, contingency planning, and practical automation Building
Regalia Division Purpose-built methods and uncommon tools for uncommon work ACTIVE
FuzzyCLAW A disciplined powerful workspace for local AI, live evidence, and purpose-built assistants. In team Testing

02 // OPERATING DOCTRINE

  • Human agency first. Systems extend judgment. They do not bury it.
  • Local by default. Models, memory, archives, and logs stay close unless the operator shares them.
  • Operator control. Users can inspect, route, archive, merge, export, and delete their intelligence.
  • No accidental disclosure. Sensitive context moves only with clear intent.
  • Repairable systems. When something breaks, the operator can see where, why, and what changed.
  • Ownership and accountability. No outside ownership without accountability. | Frontier methods and materials shaped into indispensable tools. |

03 // BUILT FOR

The end goal is an awesome future we build the systems to get there.

04 // FRONTIER STANDARD

Using Multi discipline scientific research and AI to Pave the road into the ideal future.

We are building the tools that keep people capable and competetive when the map changes: simple where possible, unconventional where useful, serious always.

Those who fear freedom are tending to cages, Selling the Locks or comfortable with the food and watching others suffer. Be brash in your freedom. There's a file in the cake.

Website · Code · Collections