Insights from FLYR Hospitality integrates AI directly into your analytics workflows to simplify how you build, explore, and communicate insights. It's not just a technical tool. It's a problem-solving partner designed to support how you already work, whether you prefer structured analysis in workbooks, open exploration in chat, or guided review through dashboards.

AI in Insights is intentionally flexible. Different users engage with data in different ways, and AI is embedded where it adds the most value rather than forcing a single workflow.

AI is embedded across several parts of your workflow, helping you build analysis, explore data flexibly, generate calculations, create visualizations, and summarize findings. Together, these capabilities reduce time spent on manual work and make it easier to move from data to decisions.

Today, AI in Insights supports seven complementary ways of working:

Each supports a different stage of analysis and different working styles, and they are designed to work together.

The Workbook Agent turns a plain-language request into a query inside the workbook experience. It's especially valuable for users who want visibility into the queries being built and the ability to directly adjust fields, filters, and calculations.

Whether you're responding to a spontaneous question or designing a multi-tab dashboard, it helps you build the query you need without assembling every field by hand.

Use everyday language to start your query:

The AI selects the right metrics, filters, and dimensions to generate your results—no need to know field names or query syntax.

You can iteratively refine your query as your needs evolve:

Behind the scenes, the AI dynamically updates your query logic—so you stay focused on the question, not the structure.

The Workbook Agent can also generate calculations during the conversation. You can request totals, variances, ratios, or shares on the fly, without manually writing formulas.

Examples include:

These calculations are built directly into your query results, so you can dig deeper without leaving the conversation. Review them before you rely on them.

​See best practices for working with the Workbook Agent.

Each interaction creates a workbook tab, forming the foundation for your dashboard visualizations. Once your query is ready, you can immediately visualize it as a table, bar chart, or another chart type. You can also request a specific format right in your prompt:

For guidance on generating or changing charts with the Workbook Agent, see How to Use the Workbook Agent to Shape Chart Visualizations.

The AI Summary visualization turns the results of one workbook tab into one short written explanation. You can place it on a dashboard like any other tile. It is not a conversation or a dashboard-wide analysis. Instead of scanning rows or interpreting charts, readers can quickly see what the data is showing.

This is especially useful when:

A summary typically covers the headline change, key trends, and anything that stands out. Like any visualization, it updates as filters or data change. On a very large result it may not read every row.

Use the Additional Context input to guide the focus and tone of the summary. Keep it short — one or two things to look at, plus any context the data can't show:

For setup, prompts, and troubleshooting, see Getting Useful Output from AI Summaries on Dashboards.

The Agent in Chat supports open-ended exploration, hypothesis testing, and broader analytical conversations. Many users prefer starting here because it feels natural, flexible, and fast for ad-hoc questions.

You can:

The AI can use uploaded documents alongside Insights data to support richer reasoning, such as incorporating marketing activity, operational plans, assumptions, or external performance data.

For pick-up, you might ask:

Recent enhancements also make it easier to move from Agent chat into structured work. You can select queries and visualizations created in chat and use them as the foundation for building a dashboard, making it a seamless path from exploration to sharing and monitoring.

Some users naturally gravitate toward this workflow because it matches how they think and explore. Others prefer the structured visibility of workbooks. Both paths are fully supported.

Use a spreadsheet tab when the job benefits from an Excel-like grid close to live Insights data. A spreadsheet tab is part of a workbook. It can connect to workbook queries, apply formulas and formatting, hold small manual inputs, and appear as a tile on a dashboard. Values you type into the spreadsheet stay inside it; they do not become queryable Insights data. For setup and examples, see Spreadsheet Tabs in Insights.

Use a dashboard when a team should return to the same curated view. A dashboard presents selected workbook analyses as tiles, with a deliberate layout and controls for the people reviewing it. That consistency is its strength: agreed measures and visuals stay together, and a recurring review is performed the same way each time.

A workbook holds the editable analysis behind a dashboard, and the dashboard presents the results you want people to see. Either can support recurring or one-off work. To create and configure a dashboard, see Dashboards. To design one that readers and the Dashboard Agent can use well, see Building Better Dashboards in Insights.

The Dashboard Agent brings conversational analytics directly into dashboards. Instead of switching back to workbooks or restarting analysis in chat, you can ask questions in the context of the dashboard you're reviewing.

The assistant uses the full dashboard as its starting point, working across all tiles together. When a dashboard is well curated around a specific type of analysis, the AI can help you navigate what matters, connect insights across tiles, and go deeper as questions arise without losing context.

When you are viewing a dashboard, the Dashboard Agent helps you analyze it by answering questions and running additional queries. When you edit a dashboard, the same side panel can help you refine its building elements, such as tiles and visuals.

On a daily pick-up dashboard, for example, you might ask:

This is especially helpful when:

Just as you can in Chat, you can upload supporting documents into the Dashboard Agent. This allows the AI to combine dashboard context, Insights data, and external inputs in a single conversation. For example, upload a marketing calendar and ask whether any campaign dates line up with an increase in pick-up.

The Dashboard Agent complements both the Workbook Agent and Chat by supporting guided exploration on top of curated analysis.

For design guidance, see Building Better Dashboards in Insights.

Use a Skill when you recognize a type of problem and want to save the method for investigating it. You choose when to start the Skill, provide or adjust its inputs, and stay in the conversation as the evidence changes the next step.

For example, an Investigate a Pickup Change Skill could ask which property and stay dates need review, validate the comparison period and revenue basis, and follow the segment, channel, or cancellation pattern behind the movement. It could finish with a supported outcome, including a conclusion that no action is needed.

Keep a one-off investigation in Chat. Save a Skill when you intend to repeat a method and choose when to run it. A Skill does not run on a schedule by itself, but a Routine can run one for you.

For setup steps, see Skills in Insights, and for templates you can copy, see Skill Examples and Templates in Insights.

Use a scheduled delivery when an existing dashboard, tile, workbook result, or formatted file is what recipients need by email. Insights refreshes and sends the content you built; it does not write a new interpretation.

Use an Alert when a condition should decide whether that existing content is sent. For example, an Alert could email an existing table of high-occupancy dates only when at least one future stay date has Occupancy of 97% or higher.

Choose these when recipients need the configured table or chart, or a PNG, PDF, XLSX, or CSV file, rather than a newly written analysis.

For steps, file formats, and options, see Content Sharing Externally from Insights: Schedules & Alerts.

Use a Routine when Insights should perform a defined analysis without someone starting a conversation, and email a fresh written response. Because nobody is present to answer questions, a Routine's property, dates, measures, and no-data behavior need to be decided in advance.

A Routine can also run a Skill you already have, so a method you trust arrives by email on a schedule. A Skill that someone has shared with you works too.

Routines and Alerts can both send an email when a condition is met. An Alert emails content you already built. A Routine writes a new analysis. You can keep an Alert and add a Routine when you want the explanation alongside it.

For an overview and setup steps, see Routines in Insights.

For wording you can copy, see Routine Examples and Templates in Insights.

To run a Skill on a schedule, see Run a Skill Automatically with a Routine.

Connections are useful when the options do different jobs. They are not a required sequence.

Use the lightest option, or combination, that completes the job without duplicating the same output.

AI in Insights supports a wide range of working styles and roles:

Whether you prefer structured workbooks, conversational exploration, or dashboard-driven review, AI supports how you work rather than forcing a single approach.

AI in Insights is designed to make analysis easier, clearer, and more adaptable to how teams actually work. If you need support or want personalized guidance, our Advisory team is always happy to help.