Version 3.36.0 20th March 2025
- Federico Delucchi
We are excited to introduce Eureka AI agents, new Action Plans and more!
Eureka AI Agents → Available only for Admin roles
We’re introducing an Anomaly Detection Agent to help you monitor feedback volumes and quickly identify potential issues affecting data collection. This ensures that you can detect and resolve anomalies before they impact your insights.
An AI agent is a smart system that monitors, detects patterns, and provides recommendations based on data. In this case, our Anomaly Detection Agent automatically tracks feedback trends and alerts you if something unusual happens—helping you stay ahead of potential issues without manual effort.
What’s New?
✅ Automated Volume Monitoring → Calculation
The agent will analyze feedback volume trends on a weekly basis.
If a sudden change in volume is detected, the agent will generate a notification with actionable insights.
We check the trends of the last 8 weeks and compare it to last week
Example:
Agent invocated the 03/03/2025
Week-8 to Week-2 → 06/01/2025 to 12/01/2025 TO 17/02/2025 to 23/03/2025
Week-1 → 24/02/2025 to 02/03/2025
If the agent identifies any change in the volume trend will trigger a recommendation with explanation if the volume change is related to pushes, imports or response rate.
✅ Smart Suggestions for Issue Resolution
Survey Issue → Redirects you to the Survey page to analyse campaigns related.
✅ Agent Notification System
You will see an alert when an anomaly is detected:
📢 "Agent Notification: We have detected an anomaly on the following source."
An email will be also send.
Go to Suggestion → Takes you to the Agents Recommendation Page to review insights.
Accept → Redirects you to the affected survey.
Reject → Strikes through the recommendation and removes the alert.
✅ Agent Management Dashboard
On the Settings Page, manage agents with:
Agent Name & Type
Invocation Frequency & Last Triggered Date
Description of the agent
Enable/Disable Toggle to activate or deactivate the Agent.
In the next iterations we will add Agents that will:
Track Satisfaction ratio on favourite attributes to detect values that are outliers and recommend a custom report to understand the anomaly.
Recommend topics to track based on your insights
How It Works
1️⃣ Automatic Monitoring – Every Monday, the agent analyzes the last 7 days of feedback data.
2️⃣ Anomaly Detection – If a significant change is found compared to the previous week, the agent identifies the affected survey.
3️⃣ Actionable Notifications – You receive a real-time alert with direct links to resolve the issue.
4️⃣ Easy Management – Enable/disable the agent in Settings..
New Action Plans
Action Plans currently provide key insights using metrics like Financial Impact and Relevance Score, but they lack one critical factor: Effort—the resources, time, and energy needed to implement an improvement. Without this, you can struggle to prioritize actions effectively.
🚀 What’s New?
1️⃣ Introducing the Effort Metric
📌 New Effort Slider (1-5 scale):
Each improvement in an Action Plan will have an Effort score (1 to 5) to indicate complexity.
The score considers cost, time, and energy required for implementation.
You can manually adjust the effort level.
📌 Eureka AI Effort Estimation:
When generating an improvement, Eureka AI will provide an estimated effort score based on cost, complexity, and required resources.
The AI-suggested score will pre-fill the effort gauge, allowing you to fine-tune it.
2️⃣ UI Redesign for Action Plans & Reports - Key Metrics Displayed:
Compact View (Title Section)
Financial Impact:
🔴 Red Badge – Revenue at risk.
🟢 Green Badge – Revenue opportunity.
Relevance Score
Expanded View (Detailed Breakdown per Improvement)
✅ Relevance Score – How important is this improvement?
💰 Financial Impact – How much money is at risk or gained?
🔧 Effort Metric (1-5 scale) – How hard is this to implement?
📊 Sentiment Breakdown – What is the emotional tone behind related feedback?
📌 Top 3 Attributes – Key factors linked to this improvement.
💬 Related Responses – Direct answers tied to the issue.
3️⃣ Editable Texts
All descriptions (except titles) will be editable.
4️⃣ Enhancements for Report Component
📊 Unified Experience: The same redesigned UI will apply to Reports.
📄 One Improvement per Page: Helps you focus on one action at a time.
Financial Impact in Topics Summarisation
In previous releases, financial impacts were displayed within action plans, but you had no visual cue to identify the financial impact of topics before creating action plans.
To address this, we're introducing a financial impact display in topic summaries to ensure you can easily spot revenue-driving opportunities or risks as soon as they analyze the feedback.
🚀 What’s New?
✅ Visual Financial Impact Indicators
Bolded text in Eureka AI topic summaries to highlight financial implications (both positive and negative).
Quickly see the revenue at risk or opportunity based on customer feedback in the summary.
✅ Immediate Insight on Financial Metrics
For each topic, the system will clearly state whether positive feedback represents revenue to keep nurturing, or negative feedback indicates revenue at risk requiring action.
Text will read:
"These negative points mentioned by customers represent €12,000 of revenue, which we recommend taking action on."
"These positive points praised by customers represent €16,000 of revenue, reinforcing the recommendation to continue in this direction."
"These positive points praised by customers represent €16,000 in revenue, while the negative points represent €12,000, highlighting the need for action."
🔑 How It Works
1️⃣ When topic summary is generated, the platform automatically calculates the financial impact of positive and negative feedback.
2️⃣ The financial impact is displayed in bold within the topic summary, allowing you to immediately assess the potential revenue implications.
3️⃣ Positive and negative feedback will be separately highlighted with the corresponding revenue amounts, based on customer sentiment
Topics Trends overtime by Percentage
Currently, when you are analyzing Text Analysis graphs you can only see trends in volume. However, volume alone doesn't provide enough insight—it doesn’t show whether a topic is becoming more or less significant relative to all feedback over time.
This feature allows you to:
✅ See the true impact of topics over time—not just raw counts but how often they appear relative to total feedback.
✅ Track improvements more accurately—confirm whether issue resolution efforts are actually working.
✅ Make data-driven decisions—prioritize topics based on their real influence rather than just absolute mentions.
How It Works
A radio button will be added to switch between Volume (default) and Percentage (%) on the Y-axis.
In Percentage mode:
Each topic’s percentage is calculated as:
(Mentions of the topic / Total responses) × 100Example: If "X topic" had 12 mentions out of 100 total responses in January, the graph will show 12% instead of just "12 mentions."
All topics will be considered, not just the top 10.
By enabling percentage-based insights, you will gain a clearer, more actionable view of how key topics evolve over time, improving their ability to respond effectively.
Feedier Team