> For the complete documentation index, see [llms.txt](https://doc.batch.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.batch.com/getting-started/features/customer-engagement-platform/batch-ai/batch-ai-assist.md). # Batch AI Assist {% hint style="info" %} Batch AI Assist features are subject to token consumption limitations. Reach out to your CSM or Account Manager to learn more. {% endhint %} Batch CEP comes with Batch AI Assist, a rich set of AI Assistants that improve CRM managers' productivity. Among them: ### Conversational Agent {% hint style="info" %} Conversational Agent is an **Alpha** feature that is being progressively rolled out to a limited number of customers. Reach out to your CSM if you're interested. {% endhint %} Conversational Agent is a conversational experience built into the Batch dashboard. It answers questions about the project you have open, using the same data and permissions as the rest of the dashboard. It can read performance data, find orchestrations, explain targeting, use the Batch documentation, and create orchestrations for you. It never saves or launches anything. For example, you can: 1. Read performance (this week vs last week, Android vs iOS, one orchestration or the whole project). 2. Find an orchestration by what it does, even if you don't remember its name. 3. Read its targeting and, on a journey, its branch conditions. 4. Draft a follow-up push, SMS, trigger journey or in-app message, review the proposal, and apply it to the form. **Where to find it** Open it from the top bar with **Open Conversational Agent**. The agent is not a modal: you can keep navigating while it works, and it follows the page you are on. Different display modes are available: * Use **Bubble** to keep the page visible, or **Side panel** for the default overlay. * Use **Pin beside the page** to dock the agent and shrink the page, or **Fullscreen** for a focused conversation. **Context awareness** The agent knows which page you are on. Based on it, it suggests relevant questions and actions. It can also suggest navigating to another page to complete a task. For now, past conversations are not memorized. **What you can ask** **Performance** Project-level and orchestration-level stats, with comparison windows and platform filters. For example: * *What is my Android push performance this week compared to last week?* * *On \[campaign], compare Android and iOS this week.* * *How are my opt-in, reachability and monthly active profiles trending?* Answers usually come with KPI tiles and a sentence interpreting the change. The agent works on aggregated profile data (opt-in distribution, reachability, monthly active profiles). It does not open an individual profile. **Search** * **By name, type, state or channel**, like the orchestration list. * **By meaning:** *Show me orchestrations that re-engage users*. Works on campaigns, recurring automations and trigger automations, including email content. Campaigns created through the API are not covered. **Orchestration flows** The agent fully understands orchestration flows: targeting, branches, and message content, except HTML email content for now. It helps you explain, compare, and suggest improvements. **Batch documentation and knowledge base** Product questions are answered using doc.batch.com and a continually improved knowledge base covering best practices and orchestration-management nuances. Ask for a citation: a good answer names the page and gives its link. **Draft a campaign or automation** You need the same rights required to create or edit orchestrations in the dashboard. Campaigns, Recurring Automations, Trigger Automations, in-app sendings, and mobile landings can be created or edited. **Data & privacy** * The agent uses your session and your permissions: it only sees what you can see in the dashboard. * It works on aggregated data and does not process individual profile data. ### AI Home A 100% AI-powered homepage that analyzes the last 7 days of sends (rolling UTC window, aggregated data only — no personal data) and turns them into scannable, actionable insights. The AI Home is built from six blocks: * **Morning Brew** — a narrative overview of your project's global health and weekly trend, rendered as metric cards (channel, value, trend, one-line insight). It reports the big picture only; it does not rank or alert on individual orchestrations. * **Campaigns** — a *Best of* and a *Flops* ranking of the week's campaigns, based on absolute performance (minimum 100 messages sent to qualify). * **Automations** — one *Trending* automation card plus an *Alerts* table, prioritized most-important-first (p0/p1/p2: engagement drop, deliverability issue, or a RUNNING automation that has stopped sending). * **Opportunities** — concrete, recommended next steps to act on. * **Ask Home** — natural-language Q\&A over your project's data. * **Orchestration preview** — clicking any orchestration or alert opens its full analytics in a side sheet, on top of the Home (no navigation, no URL change). ### AI Targeting Insights An AI expert embedded in the Query Builder that analyzes and validates your targeting logic before launch. Returns: a plain-language summary of who is being targeted, logical warnings (zero audience, mutually exclusive conditions, over-restriction), and structural suggestions to simplify or clean up your query. [Learn more](https://doc.batch.com/release-notes/release-notes/march-26-2026-ia-targeting-insights) ### AI Translation One-click AI translation of your content into one or more target languages, directly within the multi-language module. [Learn more](https://doc.batch.com/release-notes/release-notes/january-5-2026-automated-ai-translation-push-and-sms) about the SMS and Push translation. [Learn more](https://doc.batch.com/release-notes/release-notes/june-18-2026-automated-ai-translation-email) about the Email translation. ### AI A/B Test Variant Generator - Push & Email AI-powered generation of A/B test variants for Push notifications and email subject lines. The AI uses your existing variants as context to suggest the next one (B, C, D). [Learn more](https://doc.batch.com/release-notes/release-notes/october-13-2025-generate-a-b-test-variants-with-ai) ### AI Email Subject Generator Generates an email subject line from your message body using AI, directly inside the Email Composer. [Learn more](https://doc.batch.com/release-notes/release-notes/january-12-2026-ai-email-subject-generator) ### Email Quality Assistant An on-demand quality analysis of your email, embedded in the Email Builder. Click **Analyze** in the toolbar to check the content, the layout and accessibility, and the deliverability of the email version you are editing. Findings are returned ranked by severity, with a one-click fix where a safe rewrite exists, and an overall verdict on whether the email is ready to send. [Learn more about the Email Quality Assistant](https://doc.batch.com/getting-started/features/customer-engagement-platform/message/email#email-quality-assistant) ### Intelligent Warm-up An intelligent warm-up mode in Recurring Automations for gradually ramping up email sending volumes. Configurable (initial volume, increase percentage, engagement criterion) with manual intervention allowed at any point during the process. [Learn more](https://doc.batch.com/release-notes/release-notes/november-24-2025-intelligent-warm-up) ### AI Smart Naming — Automation Builder Automatically generates descriptive labels for Message and Yes/No Split steps in the Automation Builder, based on each step's actual content and configuration. Triggers automatically when closing the side-sheet, or manually via the sparkle icon. [Learn more](https://doc.batch.com/release-notes/release-notes/february-11-2026-ai-smart-naming-for-automation-buider) ### AI Profile Data Alerting Get proactive alerts on potential issues in your profile data (e.g. inactive attributes, duplicates, or inconsistent sources). Quickly review, ignore, or resolve recommendations to keep your data clean and reliable. [Learn more](https://doc.batch.com/release-notes/release-notes/march-31-2026-ai-profile-data-alerting) ### AI Profile Data Renaming Automatically get smart naming suggestions for new attributes to improve clarity and consistency. Review and confirm names before making them available in your orchestrations. [Learn more](https://doc.batch.com/release-notes/release-notes/march-31-2026-ai-profile-data-renaming) --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://doc.batch.com/getting-started/features/customer-engagement-platform/batch-ai/batch-ai-assist.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.