N17Q
A trace, replay, and policy system for testing long-running agent workflows without repeating their external consequences.
A trace, replay, and policy system for testing long-running agent workflows without repeating their external consequences.
An evidence-backed document workflow where model output remains a reviewable proposal with provenance and revision history.
Hybrid document search and grounded answers designed to refuse fluently when the collection cannot support a claim.
A staged consolidation of an over-separated service estate into a modular monolith with fewer, stronger boundaries.
A durable workflow engine that makes retries safe through stable intent, step receipts, and explicit unknown outcomes.
An asynchronous decision space built for intermittent attention, weak connections, durable handoffs, and humane recovery.
An incident narrative that joins deployments, alerts, runtime evidence, and operator annotations without confusing correlation for cause.
A collection policy and receipt system that turns privacy promises into server-enforced, testable product behavior.
A shared interface foundation organized around accessibility contracts, migration paths, and maintained design decisions.
Disposable branch previews with isolated execution, atomic routing, and content-addressed reuse of identical build work.
A file-based publishing system that replaced fragile manual releases with repeatable builds, atomic deploys, and secure delivery.
An offline inventory tool whose operation log and synchronization receipts protected field work through unreliable networks.
A live event-operations dashboard that modeled freshness and continuity instead of equating an open socket with truth.
A responsive equipment catalogue that replaced one fixed canvas with content-led layout rules and a delivery budget.