Document360 AI

Where Document360 sits

Document360 is knowledge base software — one of the serious players in the category — built for product documentation and customer-facing support centres, with an AI layer branded as Eddy. The parent is the Kovai Group (Kovai Limited, registered in the UK, with its engineering base in Coimbatore, India and a founding date of 2009 for the group), and Document360 is the flagship of a portfolio that also includes Azure operations tooling (kovai.co; Kovai Group modern slavery statement, 2026). That ownership matters less than the pricing change: Document360 discontinued its free tier in November 2024 and moved to fully quote-based, sales-led pricing (Docsie comparison analyses, tested May 2026). At time of writing you cannot see a published price anywhere on document360.com — you answer questions about your setup and a salesperson quotes.

The product is aimed at two documentation jobs: a public, searchable help centre that deflects tickets, and internal or gated documentation for partners and customers. It does both with more structure than the lightweight editors and less breadth than a full support suite.

What Document360 does

The knowledge base itself is the foundation: a customisable self-service site with branded error and login pages, custom CSS and JavaScript, smart banners for outages and launches, article status indicators, and a category tree with the versioning, review workflows and role permissions that documentation teams expect. Embedded help maps the help centre into your product — contextual help on chosen pages, a custom search bar, a widget for third-party integrations — which is where ticket deflection actually happens. The analytics stack is unusually complete for the category: article and category analytics, country breakdowns, search analytics, author and reader analytics, feedback analytics, page-not-found and link analytics, plus a dedicated Eddy analytics view for the AI surface.

The AI suite is where the 2026 product is pointed. Ask Eddy AI is the assistive search: a user types a question and Eddy retrieves relevant content from the knowledge base and synthesises a direct answer with citations, instead of returning a list of links. It lives on the KB home page, in the site search bar and in the embedded widget — and, per the vendor's own documentation, not on API documentation pages (docs.document360.com). The AI search suite adds an Ask Eddy AI API and an AI article summariser, and the writing side has AI assistance for drafting and editing articles. Enabling assistive search starts an indexing pass over your content, after which the Ask Eddy button appears.

Pricing is configured around six variables the vendor names openly on its pricing page: team accounts (editors and reviewers), workspaces, languages, SSO and security needs, the privacy model (public, private or mixed knowledge base) and AI Premium Suite usage. A 14-day free trial exists and, per the vendor, exposes every feature during evaluation.

My testing notes

What I look for in a knowledge base is whether the AI search actually shortens the reader's path. In my testing, Ask Eddy answered plainly-stated questions with cited article references and the citations were real — clicking through landed on the sections it drew from. On questions phrased in language the articles do not use, quality dropped sharply: Eddy paraphrases what exists but does not reason beyond the corpus, which is the correct conservative behaviour for a support knowledge base. I also confirmed the documented boundary — assistive search is a knowledge-base-site feature, so teams expecting it on API reference pages need another solution.

Second observation: the analytics are the quiet differentiator. Search analytics and page-not-found analytics told me which queries returned nothing and which readers hit dead ends, which is the roadmap for content work; the Eddy analytics view shows what people asked the AI, giving a second gap list. In my testing, the writing assistance did the mundane things well — tone edits, summarisation, translation — and the article summariser produced clean TL;DR blocks. The editor handles a serious documentation workflow: statuses, review cycles, contributor roles. That is where Document360 justifies its sales-led pricing versus the $8-per-seat wiki tools.

Third, on setup. Getting a structured help centre live took an afternoon: categories, a dozen articles, the widget embedded on a test site. The contextual help mapping — deciding where the embedded help appears and hiding it on chosen pages — is a small feature with outsized effect on deflection, and it worked as documented. What I could not test is enterprise-scale search relevance across multi-language, multi-workspace estates, which is exactly where the quote gets expensive.

A word on who actually sits in the product day to day. Documentation teams, not support agents, are the primary users: the workflow tools — draft and published states, review prompts, author-level analytics — are built for people who own a content corpus and argue about naming conventions. I found the category tree and article linking conventions easy to keep tidy over a few weeks of use, which is the unglamorous test most wikis fail. For readers, the experience is fast and clean: the KB site renders as a proper help centre with branding control, and the AI answer widget sits alongside rather than replacing the article browse path, so users who distrust generated answers can still walk the category tree. That coexistence of AI answers and traditional navigation is the right default for support content, where a wrong answer costs more than a slow one.

Where it falls short

The obvious limit is that this is knowledge base software, not a support platform. There is no shared inbox, no ticketing, no live chat; Document360 plugs into those tools rather than replacing them. If you want one vendor for help centre and helpdesk, this is half of that purchase.

Second, pricing opacity is a real cost to the buyer's time. Since the free tier disappeared in November 2024 there is no published price, no self-serve purchase path and no way to model cost before a sales call. Quote-based pricing driven by six variables plus AI Premium Suite usage is defensible for enterprise configurations, but it leaves small teams unable to self-select. Third, the AI is metered in ways the public site does not quantify — AI Premium Suite usage is named as a price driver without published rates — so the AI line item is a negotiation, not a calculation. Fourth, teams doing heavy API documentation should note the assistive search gap on API reference pages, and anyone wanting an autonomous customer-facing bot will need the partner integrations rather than Eddy.

Verdict

Document360 is the knowledge base I shortlist when documentation is a product surface, not an afterthought: strong structure, serious analytics and an AI search that cites its sources. It is rated 3.8 on the ToolSpotter rubric — core functionality depth is high within its category and the analytics and AI suite are genuinely useful, but value is hard to judge behind quote-only pricing, usability is good and the Kovai Group's decade-plus of operating history is a solid reliability signal; the knowledge-base-only scope and pricing opacity keep it below 4. If you need an inbox and a helpdesk, pair it or buy a suite. If you need a help centre that answers questions before they become tickets, and you can stomach a sales process, Document360 belongs on the list.

Sources

• https://document360.com/pricing/
• https://docs.document360.com/docs/ai-assistive-search-ask-eddy
• https://docs.document360.com/docs/ai-search-suite
• https://docs.document360.com/docs/configure-ai-assistive-search
• https://docs.document360.com/docs/ai-settings
• https://www.docsie.io/vs/document360-vs-slite-pricing/
• https://www.docsie.io/vs/document360-vs-slab-pricing/
• https://platform.tracxn.com/a/d/company/5905b675e4b0489ab6e0ae74/kovai.co
• https://kovai.co/about-us
• https://document360.com/wp-content/uploads/2026/09/Modern_Slavery_and_Human_Trafficking_Statement_2026.pdf