THIAGO AUGUSTO LUCAS DA SILVA

Working with Foundation Model for Salt bodies projects.

• Optimization of a well drilling campaign;
• Prediction of failure in pumps used in hydraulic fracturing (Fracking) - Current.
• Carry out extensive literature searches on solutions to combinatorial optimization problems, focusing on exact solutions, metaheuristics and MLCO (Machine Learning for Combinatorial Optimization);
• Testing solvers for combinatorial problems;
• Developing an exact optimizer for well drilling campaigns, extending solutions from the Flexible Job Shop Scheduling Problem (FJSP) to the case of the Rig Scheduling Problem (RSP), inserting new restrictions that bring the problem closer to the real scenario of a campaign; This optimizer was implemented on the theory of Mixed Integer Linear Programming via Python with PySCIPOpt and the SCIP solver;
• Seek improvements to the exact optimizer: Provide the solver with relaxed solutions initially and implement an internal Branch-and-Bound method;
• Extend Machine Learning solutions applied to FJSP to RSP;
• Perform sanity tests on each implemented constraint;
• Perform computational experiments on benchmarks from the literature and instances that simulate the real campaign scenario;
• Generate clear visualizations of the optimized campaign through Gantt charts, highlighting each restriction implemented such as the start and end of the contract for each rig in the schedule, the First Oil Date for each well, the preparation/denial time required by each rig, among others.

• Delivery time calculator: Reduction in delivery time promise.
• Control tower: Monitoring's structure of o2o orders.
• Data Analysis, especially exploratory, descriptive and diagnostic.
• Viability analysis of project integration.
• Generate indicator for performance measuring.
• Build and maintenance of Analytical databases by query, cleaning, wrangling, transformation and validation of data, as the implementation of new features for Python algorithms, using tools from Google Cloud Platform (GCP) as BigQuery, Storage e Compute Engine - Jupyter Notebook, from Amazon Web Services (AWS) as VM's EC2, and a PostgreSQL database.
• Structuring, documentation and code versioning - GitLab.
• Directly align the stakeholders.
• Agile methodology: Scrum, supported by tools like Jira Cloud and Confluence Cloud.

• Research in Dynamical Systems, Ergodic Theory.
• Interrupted

• Research in Dynamical Systems, Ergodic Theory, especially on the results about a class of Linear Cocycles particularly studied by Michael Herman which is possible to find a lower bound for the maximal Lyapunov Exponent.
• Provide the proof's details of the theorem 'The formula of Ávila-Bochi-Herman'.
• Main required areas: Real Analysis, Complex Analysis, General Topology and Algebraic Topology.

• Development of computational tools as the first version of the project's graphics generator, starting with desktop applications using the Netbeans (IDE), Java (language) and the software JFreeChart. Further we used Java Web Applications, the software Primefaces and a MySQL database.
• Project: ADACA

• Search for strategies in order to facilitate the communication, interaction and learning of autistic children.
• I joined the group responsible for the management of all the data of the project.
• Development of tools which allow the project to be structured and individually analyze the progress of each child.
• Project: ADACA