> For the complete documentation index, see [llms.txt](https://ask.birdie.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ask.birdie.ai/product-updates-changelog/2026-q2.md). # 2026-Q2 Q2 2026 made it possible to talk to your customer data through any AI assistant. That was the headline, and it changed how teams access and share Customer Intelligence insights. Around that launch, the quarter also gave teams the ability to detect feedback changes automatically, compare channels side by side on a schedule, and manage quality monitoring with more control and fewer workarounds. *** ## Birdie and AI: The MCP Launch You can now ask an AI assistant (Claude, ChatGPT, Gemini, Cursor, or any tool that supports MCP) questions about your customer feedback and get answers from your actual Birdie data. Instead of logging in, setting filters, and building a view, you ask: * *"What are our top five complaints this month?"* * *"Show me NPS trends over the last six months."* * *"Find negative feedback about onboarding from enterprise customers."* * *"How many feedbacks did we receive this quarter, grouped by segment?"* * *"Compare satisfaction scores across our main support channels."* The assistant understands your taxonomy, filters, segments, and data sources. Every query runs against your organization's data only, scoped automatically to your account. You sign in with your existing Birdie credentials, so there is nothing new to set up on the access side. This means Customer Intelligence analysis can happen wherever your team already works, inside a chat window, a coding environment, or a document editor, rather than requiring everyone to know their way around the Birdie interface. [Learn more about the Birdie MCP Server](https://github.com/birdie-ai/gitbook/tree/main/integrations-and-data-ingestion/birdie-mcp-server.md) *** ## More in Q2 ### Comparison Digests Teams that monitor feedback across multiple channels can now receive a regular side by side breakdown automatically, without running the comparison manually each time. A new digest type, **Source Breakdown**, lets you define named source groups (for example, "WhatsApp" and "Email Support") and schedule a delivery that shows how volume and themes differ across each group. It arrives in your inbox or Slack channel at whatever cadence works for your team. The result is that differences between channels surface on a routine basis, not only when someone thinks to check. [Learn more about Digests](/admin-and-settings/notifications/digests.md) ### Trends and Anomalies Teams no longer need to check dashboards to find out when something unusual is happening in their feedback. Birdie does it for them. Two detection modes are available. **Spike Detection** alerts you when feedback volume on a source rises abnormally compared to baseline, so you hear about a support surge or product incident at 8:00 AM rather than in the weekly review. **New Pattern Detection** alerts you when the volume of an entity shifts into a new sustained pattern. For example, if an area that used to receive 100 feedbacks a week starts consistently receiving 120, that change in pattern is what triggers the alert. You control the sensitivity, where the alert is delivered, and the time zone. Birdie runs the analysis daily. [Learn more about Anomalies](/admin-and-settings/notifications/anomalies.md) ### Taxonomy Catalog FI managers can now see, edit, and manage the full evaluation framework from one place instead of navigating across multiple pages. The catalog organizes all criteria into three tabs: **Active** (live criteria, editable in batch so you can make multiple changes before committing), **Collections** (criteria grouped by team or topic), and **Pending** (criteria suggested by AI, waiting for review before they go live). You can calibrate, duplicate, and adjust AI instructions for any criterion directly from the catalog, which means tuning the evaluation model no longer requires jumping between settings pages. [Learn more about Taxonomy](/admin-and-settings/taxonomy.md) ### Connections Admins can now manage all data source credentials in one place and reuse them across integrations, rather than entering the same credentials again every time a new pipeline is created. The Connections page shows every stored credential, whether it is active or expiring, and which integrations depend on it. When a token is about to lapse, you can see exactly what will be affected and update it once. ### Disputes Agents and supervisors can now formally challenge a quality evaluation they believe is wrong, with a record that follows the process from submission to resolution. A dispute includes a written justification for each contested criterion, a reviewer workflow where decisions are made criterion by criterion, and an automatic correction to the quality record when a dispute is accepted. Nothing changes silently. For FI managers, the bigger value is in the analytics view. Criteria that collect many disputes and have a high acceptance rate are telling you that the AI instructions need to be revised. Birdie surfaces this pattern directly, so recalibration decisions are based on evidence rather than intuition. [Learn more about Disputes](/frontline-intelligence/disputes.md) --- # 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 dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://ask.birdie.ai/product-updates-changelog/2026-q2.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. 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.