.png)
1. Final Interviews
This week, with it being the final week of the challenge and most of the knowledge base being settled, it was time to identify a few possible action plans for the company and to review the solution/implementation with both Diogo and Tiago.
As presented before, company action plans are suggested by the expert system when a given relative frequency of clients show dissatisfaction with a given aspect of the company's services (e.g. 30% of the clients are dissatisfied with the payment methods).
For the given example, an appropriate company action plan would be:
Dissatisfaction Cause(s) | Analysed Metric | Action Plan |
Frequent abandonment when choosing the method of payment (e.g. 30% of the clients) | Continuity Rate - from Payment to Review | 1. Validate if the provided payment methods are aligned with the current needs of the company's market and target-clients 2. Add new payment methods to the platform 3. Send a communication to clients telling them that new payment methods were added to the store |
Overall, both experts were very satisfied with the final solution, agreeing that with further refinements and extensions, the expert system could very well surpass their individual performances, both in terms of efficiency and time.
A few of the proposed refinements include:
Improve explanations and the provided information;
Collect more diverse client data;
Compare the company's results with the competition;
Allow the company to configure its own client metrics/data.
2. Composing the Web Application
With most of the implementation complete, this week it was time to finish what was left to do (e.g. toggle between both expert systems in the single page application) and implement identified refinements in each one of the modules, in order to deliver the most polished and complete solution possible by the end of the week.
Here's a short look at the final web application, where the single page application is quickly explored and interacted with.
It is worth noting that the team had to split its time this week between the solution implementation and the writing of the reports/articles for each course unit, which ultimately made the team leave a few refinements and adjustments to be done.
3. Week Retrospective
In retrospective, the final week of the project was a rather stressful one, with the team wanting to deliver the best possible solution, but with time fighting against it.
In the end, the team believes that it delivered the best possible product with the remaining time, with further improvements and extensions being quickly identified and taken note of, so people in the future can pick up where this solution was left off.
With this said, next week the team will have the opportunity to showcase what was done in the past seven weeks, both to the professors in a technical demonstration, and to the AI community at Demo Day. The team can't wait to see what will come out of them.
Comments