--- title: "Introducing ADA: Your Rovo Agent for ActionableAgile® Analytics" canonical: "https://support.55degrees.se/space/SP/4261412865/Introducing%20ADA%3A%20Your%20Rovo%20Agent%20for%20ActionableAgile%C2%AE%20Analytics" format: markdown --- [Jira Cloud] [8.3.0+] [Beta] > ℹ️ We updated our DPA prior to Ada’s release to give information about what’s involved when you utilise this feature: [https://www.55degrees.se/legal/customer-agreements/atlassian-cloud-products](https://www.55degrees.se/legal/customer-agreements/atlassian-cloud-products) > ℹ️ That’s a *big* data set you’ve got there! We’ve identified an issue of large data sets (5,000+) timing out before Ada can analyse them due the API gets throttled by Atlassian. > ℹ️ > ℹ️ We’re working on implementing a cap to stop this but please be aware to only analyse the data you really need to by refining your data sets. ### Meet Ada: A Flow and Delivery Analysis Assistant Ada is your Rovo Agent designed to explore delivery flow and work patterns using your data and ActionableAgile® Analytics. Think of Ada as a focused partner for Cycle Time, Throughput, Aging WIP, bottlenecks, and probabilistic forecasting - not a general-purpose chatbot. #### So, What Can Ada Do? - Answer practical questions like: *How predictable is our work? Where are items getting stuck?* *What looks risky or old? What should we focus on next? What are realistic delivery expectations?* - Work with **both** saved ActionableAgile® Analytics datasets and on-demand Jira Board data. - Explain results in **plain** *language*, **highlights** *risks*, and **suggests** *next steps*. #### Who Can Ada Help? Well, everybody really. Ada is designed to be utilised by anyone who uses ActionableAgile® Analytics, so they can make the best out of their metrics. For specifics, we’ve briefly listed some common positions and how they might use Ada below. - **Executives** and **Product Managers** by giving them clear info about delivery status, risks, and how confident we are in forecasts. - It supports **Delivery** and **Flow Managers** with useful stats on Cycle Time, Throughput, bottlenecks, Aging Work in Progress (WIP), and forecasting. - **Engineering Managers** and **Team Leads** can see which items are at risk, which tasks are blocked, and what needs to be done next. - **Agile Coaches** and **Scrum Masters** can look at flow patterns, spot chances to improve, and use the theory behind it all. - **Developers** get advice on which tasks to finish, unblock, or start. - Lastly, **Jira Admins** and **Operations Staff** get help with board-based analysis, checking workloads, filtering, and making sure Views are correct. --- ### Enabling Ada A **Jira Admin** can independently enable AI capabilities which will in turn give you access to Ada. Navigate to **Jira Admin Settings** → **Apps** → **ActionableAgile® Analytics**, the option to **Enable AI Capabilities** is located under **Available Configurations**. Both or neither, one or the other, it’s up to you to choose which features you want to have enabled. We recommend that both are enabled for the fullest experience but either feature works regardless of the other. - If you want users to be able to analyse metrics but not receive coaching around them - only enable AI Insights. - On the other hand, if you’d like general Agile advice about hypothetical contexts - only enable AI Coaching. > ℹ️ We train Ada’s coaching using the progenitor of the app itself - Actionable Agile: Metrics for Predictability by Daniel Vicanti. ### Getting Started #### Where is Ada? As you can analyse both your ActionableAgile® Analytics data sets and your native Jira Boards, we thought it best that you should be able to access Ada in both places. ##### In the App > ℹ️ **Context** is *key* - when using Ada in the app, we recommend loading the data set you want to analyse before booting up AI, so you can compare results in real time. ##### In a Jira Board > ℹ️ Can’t find Ada? There’s two possibilities. > ℹ️ > ℹ️ 1. It’s likely it **hasn’t been enabled** by your Jira Admin, so make sure to check with them that they’ve toggled on AI Capabilities in the Jira Admin Settings. > ℹ️ 2. You’re **not on the latest version** of ActionableAgile® Analytics - only **8.3.0** and later have access to Ada, so **[upgrade now](https://support.55degrees.se/space/SP/4243324929/Admin+Upgrade+Guide)**! #### Talking Points Now that you’re confident Ada can help and where to find it, what’s the best way to get started? Well - after extensive testing on our side - we suggest beginning with a saved dataset or quickly spin up analysis from a Jira Board. Ada guides you to the next sensible step and avoids assumptions. When Ada first boots up, it’ll politely introduce itself and tell you what it’s all about. The intro might vary slightly whether you have ActionableAgile® Analytics loaded with or without a data set or whether you’re accessing Ada directly from your Jira Board. ##### Example Prompting | **Outcome** | **Prompt** | **Must** | | --- | --- | --- | | Ada introduces itself | “*What can you help me with?*” or “*How should I get started?*” | | | List available data sets | “*List my datasets*” | Provide part of a dataset name to narrow results. | | Find data sets that contain a specific issue | “*Which data sets contain ABC-1234?*” | Mention a Work Item Key or ID. | | Load a specific data set | “*Load data set Quarterly Review.*” | Name the data set. | | Load a data set with a specific View | “*Load data set Quarterly Review using the view Q3*” | Name the data set and the View. | | Load and analyse a Jira Board | “*Load board data for Team Phoenix.*” or “*Analyse board 123 for the last 90 days*” | Specify the Jira Board name or ID.
Optionally mention a date/look-back, extra JQL filtering for more specific results. | | Apply a View to a data set | “*Apply the Expedite view to this dataset.*” or “*Switch to the Default view.*” | Specify the View. | | Update time-frame and fields filters | “*Show only the last 30 days*” “*Filter to High and Critical priority items*” “*Clear all filters*” | | > ℹ️ Start broad, then narrow with views and filters. State your intent (e.g., “focus on risk” or “show aging work”) so Ada can tailor guidance and analysis. ##### Best Practices - Be **explicit**: Mention whether you want analysis (like cycle time), action (such as focus recommendations), or discovery (fields/datasets). - The early **scope** gets the worm: Load a dataset or board *before* asking analysis questions. - Stay **relevant** with Views and filters: Narrow to pre-made Views, specific attributes, or time windows for focused insights. - Never deal in **absolutes**: Forecasts are probabilistic; ask for confidence levels or likely ranges instead of certain outcomes. - **Beyond** small-talk: Ask Ada to dig deeper on the signals you care about. > ℹ️ **Common pitfalls to avoid:** > ℹ️ > ℹ️ - Asking for analysis before selecting a dataset or board. > ℹ️ - Asking for a forecast without target items or a target date. > ℹ️ - Expecting single-point certainty rather than probability-based results. --- ### Analysis & Reporting Skills Once you’re familiar with how best to prompt Ada to get you the data you want to see, we can dive a little deeper into the analysis and reporting skills on offer. > ℹ️ ***All ***skills listed below **require** the user to have **loaded a data set or Jira Board** context with Ada before the analysis, reporting or coaching can commence. #### **Cycle Time & Predictability ** - Analyse [Cycle Time](https://support.55degrees.se/space/SP/4312957014/Cycle+Time+Scatterplot) Distribution - **Purpose**: To get a good idea of *how long* work usually takes, utilising percentiles, the range of times, and any unusual cases. - **Options**: Percentiles (e.g., 50th, 85th, 95th), include blocked time, time unit. - **Prompt**: “*Explain our cycle time*” → “*What is our cycle time at the 85th percentile?*” → “*What would our cycle time be without time spent blocked?*”. - **Result**: A readable summary of typical durations, signals, variability, and notable slow items. #### **Current Risks & Flow Signals ** - Analyse [Aging Work in Progress](https://support.55degrees.se/space/SP/4300472321/Aging+Work+in+Progress) - **Purpose**: Spot any in-progress tasks that seem *old*, *risky*, or *stuck* compared to previous work. - **Prompt**: “*Show our aging work in progress*” → “*What work looks stuck right now?*” - **Result**: Items and patterns that merit attention, with rationale. - Analyse [Cumulative Flow Diagram (CFD)](https://support.55degrees.se/space/SP/4343332869/Cumulative+Flow+Diagram) - **Purpose**: See distribution across workflow stages and identify bottlenecks or imbalance. - **Prompt**: “*Show cumulative flow signals” *→ “*Where is work getting stuck in the workflow*?” - **Result**: Workflow stage-level patterns, likely constraints, and impacts on flow. #### **Delivery Pace and Stability ** - Analyse [Throughput Trends](https://support.55degrees.se/space/SP/4363059215/Throughput+Run+Chart) - **Purpose**: See how much work *actually* gets done and how much the delivery can *change*. - **Options**: Daily or weekly time-frames. - **Prompt**: “*What does our throughput look like?*” → “*What are our weekly throughput trends?*” - **Result**: A clear explanation of the pace, patterns, and variability of your Throughput. #### **Forecasting and Expectations ** - Run Monte Carlo Delivery Forecast - **Purpose**: These are forecasts based on probability, showing either *[how many](https://support.55degrees.se/space/SP/4302503954/Monte+Carlo:+How+Many%3F)* items you might get by a certain date or *[when](https://support.55degrees.se/space/SP/4303257603/Monte+Carlo:+When%3F)* you might get a certain number of items . - **Inputs**: Target item count or target date; optional confidence levels and simulation runs. - **Prompt**: “*When are we likely to finish 20 items?*” → “*What is the chance of finishing 15 items by June 30th?*” - **Result**: Ranges and probabilities - not single-point promises. #### **Actionable Guidance** - Get Prioritised Work Recommendations - **Purpose**: Practical next steps: what to focus on, what risks to watch out for, what needs unblocking, and what to finish before moving on to new tasks. - **Prompt**: “*What should I focus on next?*” → “*What work is most at risk?*” → “*What should we finish before starting more?*” - **Result**: A prioritised, rationale-backed list you can act on. - Generate Executive Summary [coming soon] - **Purpose**: A simple high-level overview showing how delivery is going, any risks, and what to expect next. - **Options**: Target items, target date, percentile/confidence preference. - **Prompt**: “*Create an executive summary*” → “*Summarise delivery health for leadership*” - **Result**: A clear and concise overview suitable for everyone. #### **Data Discovery and Inspection** - Get Field Insights - **Purpose**: Discover common fields and values you can use to create your questions and filters. - **Options**: Field name, value-name. - **Prompt**: “*What fields are available in this dataset?*” → “*Show me the most common values for Priority*” - **Result**: A list of fields and their most common values with counts; this is useful to check out before diving into a deeper analysis. --- #### **Coaching & Interpretation** - Ask About Agile & Flow Metrics - **Purpose**: Theory and best practices across flow metrics, Kanban, Scrum, WIP limits, blocked time, and more. - **Prompt**: “*What does a widening cycle-time spread usually suggest*?” → “*When should teams use WIP limits?*” → “*How should we interpret blocked-time signals?*” - **Result**: Concise and practical coaching to interpret signals and guide improvement. > ℹ️ Ada doesn’t jump to conclusions, it talks about chances, gives ranges, and uses phrases like “what this suggests.” Plus, it always offers clear next steps so you’re not left hanging. --- ### Troubleshooting As with all things in [Beta], not everything will be working perfectly. We encourage you to always double-check the results provided by Ada and if you encounter any issues with Ada specifically, please feel free to contact us at [support@55degrees.se](mailto:support@55degrees.se). ![Ada - Debug response.png](media://e169c023-531e-4e42-aa21-4d59b4da0c41) --- ### FAQ
Do I need a saved dataset, or can I start from a Jira board? You can choose either option. If you have a ready-to-use ActionableAgile® Analytics dataset with Views, start with that for steady analysis. If not, just ask Ada to load a Jira board for quick analysis and then narrow it down using filters.
How does Ada handle blocked time and aging WIP? When it comes to Cycle Time, you can decide whether to count the time an item is blocked. For Aging Work in Progress, Ada checks how long an item has been in progress compared to the usual Cycle Time. This helps spot any risks or delays early on.
What makes Ada’s forecasts trustworthy? Ada uses Monte Carlo simulations to show results as ranges with chances attached. It doesn’t make exact promises but shares how confident it is, so you can have realistic expectations.
Can Ada show me the raw issues behind a chart or summary? Totally, just use the prompt “Retrieve issues” with optional fields, filters, limits, and sorting to check out some typical items in scope.
What if my request is vague or missing inputs? Ada will ask some questions to clear things up and then suggest what to do next. This might include picking a dataset, applying a valid View, or giving forecast targets.