Recorded case

Training our own model: a gated research direction

A proprietary model is an option to test after retrieval, prompts and routing have measurable limits.

Reviewed 15 October 2026 · scope: planned

Recorded method and findings

The immediate asset is a curated corpus of accepted tasks, failed attempts, receipts and evaluation cases with appropriate rights and privacy controls. Accumulating conversation history is not training model weights, and a useful knowledge index is not evidence of a trained model.

Recorded method and findings

Before a training run, define the target task, holdout set, baseline, compute ceiling and success threshold. Compare improvement with simpler retrieval and routing changes. Publish the method and measured limits only if the result passes review; no trained proprietary model or training advantage is claimed here.

Practical checklist

Use only data with a clear permission and retention basis. Keep evaluation examples separate from training data. Measure accepted quality, total cost and latency before scaling.

Sources & scope

Swarm coordination implementation specification 2026-10-15
Internal retrieval source and freshness contract 2026-10-15

Keep exploring

AI orchestration: from a prompt to an accepted resultA practical design for delegation, evidence and recovery.Shared memory without shared mistakesKeep facts, hypotheses and permissions distinct across agent sessions.Engineering that can show its workTVIJO connects operators, bounded agent work and knowledge that can be checked and reused.AGENTORCH: the orchestrator has a concrete jobAGENTORCH names the orchestration role: plan, assign, observe, reconcile and accept.The task packet is the interfaceA good task describes a verifiable outcome rather than an invitation to improvise without limits.AGENTJUDGE: acceptance follows evidenceA reviewer checks the task, the artifact version and each acceptance criterion.Measure cost per accepted resultSmall task benchmarks illustrate why task class and acceptance matter more than a cheap model label.Budgets and outages belong in the state machineA reliable operator view distinguishes working, waiting, failed, blocked and accounting uncertainty.Keep VECTORBASE, evaluate retrievalUse a private source-linked index to find evidence; do not treat search results as authority.SPIDER: evidence collection and useful progressThe first implementation reads an explicit project-evidence list and produces a dated operator digest without model calls.DSHREMBOT and tmux: one operator viewAuthenticated project commands, budget visibility and explicit schedule controls connect the operator to existing execution lanes.The AI worker–Hugging Face incident: learn from controlsThe incident reports support specific engineering lessons, not a controlled ranking of our system.What the supplied AI worker screenshot actually provesThe screenshot records a safeguard-triggered switch in one session; it does not explain the exact trigger. The observation is the supplied screenshot dated 15 October 2026.A bilingual knowledge product with visible limitsPublish useful explanations first, then expand languages when the evidence and workflow are ready.Free learning, custom projects and future challengesTurn evidence-backed engineering into clear offers and a useful developer community.Trust the sender, not a public checksumBind the complete message to an explicitly trusted authority.