Ada CX

Ada CX is an AI customer experience (CX) platform built around an "AI Agent" that uses resources you provide (such as knowledge base content) to interact with end users. The documentation here focuses on its data retention policies and security controls rather than the conversational features themselves, but it’s clear the product is designed for organizations that run AI‑assisted conversations and need to manage the resulting data responsibly.

You manage Ada through a web dashboard, with roles (such as Owner and Admin) controlling who can configure security settings. There are also APIs for data compliance and data export, which you can use to programmatically delete or retrieve end‑user data and transcripts. Setup appears to be cloud‑hosted and managed by Ada; detailed onboarding or integration steps aren’t covered in the content provided.

Based on the available docs, Ada leans heavily into governance: automatic deletion of end‑user data after two years, configurable redaction of sensitive information before it’s stored, and APIs to delete or export data on demand. According to their documentation, data sent to large language model (LLM) providers is processed in memory only and not persistently stored, following Zero Data Retention principles. Admin features like configurable session timeouts, role‑based access to settings, and optional SMS‑based two‑factor authentication reinforce the enterprise focus.

What’s less clear from these pages is how agentic the AI actually is beyond handling conversations with end users. The docs don’t describe complex multi‑step workflows, tool use beyond Ada’s own APIs, or long‑term task automation, so if you’re looking for a general‑purpose autonomous agent platform, you’ll need to dig deeper into their broader product documentation. Pricing details and plan structure are not mentioned in the materials provided; you’re directed instead to contact an Ada representative, which suggests an enterprise‑oriented sales motion. If you require self‑hosted or open‑source deployment, the docs we saw don’t mention those options, so you should confirm deployment models and compliance fit directly with Ada.