Wikova
Wikova is an AI-generated, Wikipedia-style knowledge-base platform operated by Wikova PTE. LTD., a Singapore-registered company[^c1]. It is presented as a personal knowledge base powered by a team of AI agents[^c2]: a user gives it a topic, and the agents research it, structure the findings into cited wiki entries, and continue maintaining them over time. The service is available as a free iOS app and through the wikova.com web platform, and it is in public beta[^c5].
The product builds on the LLM Wiki knowledge-management pattern that Andrej Karpathy introduced in April 2026[^c4]. In this model an agent acts as a compiler: raw source material is read once at ingest time and turned into a structured, interlinked wiki, which the same agents keep up to date, instead of retrieving fragments from raw documents on every question. On Wikova this pipeline runs as a chain of agents — a Researcher, a Curator, and an Inspector — described in [[overview/what-is-wikova]].
Each wiki carries the features of a continuously maintained reference work: every claim links back to the original source[^c6], followed topics refresh daily[^c7], and the underlying architecture pairs a Next.js single-page-application frontend with a separate API backend[^c8]. The platform occupies the intersection of personal knowledge management, AI-generated content, and agent-maintained knowledge bases, competing with both note-taking applications and commercial AI knowledge-base software.