Events

In the room, in the field, with the machines running.

Featured

Sovereign AI Week

Builders, researchers, infrastructure companies, and enterprises exploring AI that can operate locally, privately, and independently.

GLOBAL LIVE STREAM

A week for people building AI they control — local models, private infrastructure, agents that do real work, edge deployments, and physical systems.

The programme is being designed around working systems rather than prediction: short field briefings, live stack walkthroughs, published Lab methods, and operator conversations about cost, reliability, security, and control.

The intended audience is builders, infrastructure operators, researchers, technical leaders, and public-interest organisations making concrete deployment decisions. Product promotion will not substitute for a method, architecture, or result that attendees can inspect.

Current state · announced. Programme, speakers, capacity, access model, and any fee are not confirmed. Registration is not open, and no third party is authorised to sell access on our behalf.

Programme formats

Field Labs

Small hands-on sessions where a stack gets built, measured, and stress-tested in the room rather than described on a slide.

SEATTLE + FIELD · IN PLANNING

Field Labs pair a working setup with the people who have to operate it. A session begins with a published question and configuration, then uses the room to install, run, measure, break, and recover the system.

The proposed format is deliberately small. Participants should leave with the method, configuration, raw observations, and failure notes—not only a demo. Candidate sessions will come from the Lab queue: the $10K local stack, a cloud-versus-local comparison, and verified agent execution.

Current state · in planning. No date, venue, capacity, equipment partner, or application process is confirmed. A Field Lab will move to scheduled only after its test method and host requirements are published.

Workshops

Working sessions on running open models, private infrastructure, and agents inside real constraints.

GLOBAL LIVE STREAM · IN PLANNING

Workshops are working sessions for one bounded skill: selecting a model against an evaluation, sizing memory, standing up a private runtime, tracing an agent, or designing an exit test for a hosted dependency.

Each workshop will publish prerequisites, the expected artifact, and a verification step in advance. Remote sessions must work without privileged access to an instructor’s environment; in-person sessions must state the hardware participants need or provide it explicitly.

Current state · in planning. Topics, instructors, dates, capacity, price, and registration are not confirmed. The first workshop will be announced only when its materials can be completed with the documented setup.

Builder Meetups

Recurring, informal, and technical — people who are actually running the stack comparing notes.

SEATTLE · IN PLANNING

Builder Meetups are intended as recurring technical exchanges for people running parts of the stack. The useful unit is a concrete observation: a model that passed or failed a workload, a configuration that reproduced, an outage, an unexpected cost, or a tool that removed a dependency.

The proposed format is informal, with brief show-and-tell slots and an operator roundtable. Sales pitches and unsourced performance claims are out of scope. Notes may be published only with the speaker’s permission and will distinguish measured results from personal experience.

Current state · in planning. Cadence, venue, first date, host, capacity, and access model are not confirmed. There is no registration list yet.

Registration is not open. Dates, venues, and partners are being confirmed.