- Open administration
A decade of recognition: the special awards to celebrate
The last analyzed period of the AOC's generative AI chatbot (from March 15 to April 14, 2025), has obtained exceptional results, equaling the best data recorded during the January evaluation.
This evaluation is based on a comprehensive review of 150 conversations, randomly selected and conducted at different times of the day, ensuring a complete and representative analysis of the chatbot's performance.
In this case, during the last review, the chatbot obtained a success rate of 91,33% (137 correct answers and 13 incorrect ones). The most consulted AOC services were the following:
- IdCAT Mobile (30,67%)
- IdCAT Certificate (16,67%)
- Enotum (15,33%)
- Effect (0.67 %)
The remaining 36% were consultations outside the scope of AOC services.
Evaluation methodology
The evaluation is carried out monthly, analyzing 150 randomly selected conversations. Each conversation is reviewed to determine whether the chatbot has correctly answered the problem raised by the citizen.
Analysis of conversations at different times of the day has allowed us to identify that the chatbot maintains constant performance, regardless of the time of day at which the interaction takes place.
Conclusion
The results for this month of March match the best data recorded in January (91,33%) and this reinforces confidence in the chatbot's ability to offer a quality service on a continuous basis and an alternative to traditional support.
Furthermore, it should be noted that during the month of February the post was published about the increase in the number of requests managed by the support service These results highlight the importance of AI in improving public services and its ability to adapt to user needs.
You can also check the following news related to the chatbot: