Open for application · MSc

"Agentification" in a Manufacturing Context

Supervised by Arlindo L. Oliveira

Development and implementation of an ecosystem of distributed intelligent agents capable of automating administrative and operational processes in a manufacturing context, ensuring interoperability between heterogeneous data sources (ERP, email, databases) and optimizing the value chain. The challenges inherent to this project include:

- Multi-source integration architecture: developing a robust connector system that ensures bidirectional data flow between agents and external sources (SAP, email servers, production logs, and internal databases).

- NLP processing agents: creating agents specialized in the automatic triage and categorization of incidents, capable of interpreting free text and routing issues to the responsible departments.

- Automated data extraction: developing intelligent extraction modules for documents (invoices, delivery notes, purchase orders, etc.) to automatically feed the management system, reducing manual errors.

- Quality Assurance (QA) optimization: implementing monitoring agents that, through data analysis, identify non-conformance patterns and suggest corrective actions.

- Agent orchestrators: creating a workflow management layer that coordinates communication between the different agents, ensuring that the output of one process correctly triggers the next action.

- Industrial validation: evaluating the impact of the solution through performance metrics (KPIs), such as reduction in incident response time, data extraction precision rates, and efficiency gains in quality control.

Cooperation with company or external entity: The work will be developed with Renova's teams.

Requisites

In addition to Renova valuing student proximity and this experience contributing enrichingly to professional preparation, the success of the project requires the student to be present at Renova's facilities on at least some days of the week.

← All dissertations