bottest.ai

bottest.ai is an AI chat bot tool. Open-source, no-code platform for testing AI chatbots. Record conversations, set baselines, rerun suites, and get pass/fail analysis plus security and adversarial tests, all self-hosted via Docker.

Bottest is an open-source testing platform for conversational AI, aimed at teams that ship chatbots and need a repeatable way to check that they still behave correctly after every change. The product wraps a browser-based recorder around three ideas: capture a real conversation, define what a good response looks like, then rerun the same scenario on demand. Its pitch is that subject matter experts stay in control of testing instead of handing every check to engineers. The homepage frames this as "test smarter not harder" and positions the tool as no-code automation for chatbot QA. The workflow is deliberately short. A user records tests and sets baselines for a conversation with a few clicks. On replay, each response is measured against the established baseline rather than against exact string matching, and success criteria for those AI-powered evaluations can be customized. After a suite run, results settle into pass, fail or mixed states, with per-test analysis explaining why something failed. Beyond functional regression, the platform covers performance tracking over time and prepackaged chatbot-specific security tests, plus automatic test case generation intended to widen coverage without hand-writing every scenario. A separate adversarial testing line targets chatbots that must resist manipulation. Deployment is the differentiator for regulated and privacy-sensitive teams. Bottest is free, hosted on GitHub, and installs with a single Docker Compose command on your own infrastructure so conversation data never leaves your environment. That fits organizations that cannot send customer transcripts to a third-party SaaS testing vendor. The trade-off is that Bottest behaves like a developer-oriented open-source project rather than a polished commercial suite. Setup, upgrades and infrastructure belong to you, and the public documentation lives primarily in the GitHub readme. Teams wanting a managed chatbot QA service with vendor support should weigh that against the self-hosted model before committing.