Writing
Publishing STRATON-LLM in public.
Technical writing on STRATON-LLM, ontology negotiation, multi-agent communication, evaluation, and system design. Each post connects a core idea to an implemented or clearly engineered system behavior.
April 5, 2026 · 8 min read
Designing Semantic Negotiation for Heterogeneous Intelligent Agents
Why STRATON-LLM treats ontology mismatch as a negotiation problem rather than a simple alignment problem, and how the end-to-end system is structured.
April 4, 2026 · 7 min read
Why Ontology Alignment Alone Is Not Enough for Real Agent Interoperability
Ontology alignment helps agents find candidate correspondences, but real interoperability needs context, confidence handling, negotiation, and post-dialogue evaluation.
April 3, 2026 · 8 min read
How Protocol-Governed Dialogue Makes LLM-Supported Negotiation Inspectable
The protocol engine is the control spine of STRATON-LLM. It turns negotiation from a free-form exchange into a replayable, rule-bound process.
April 2, 2026 · 7 min read
Evaluating Agent Agreements with Dialogue-Aware Trust Heuristics
STRATON-LLM evaluates the quality of an agreement by scoring the dialogue that produced it, not just the final answer.