Quants & Financial Markets
Simulations & Software
I’m Jose Luis Silva, AI Physicist, Ph.D., and Entrepreneur. I am passionate about Artificial Intelligence, Data-Driven Decision-Making, Modeling & Simulation. I hold a Ph.D. in Physics from Uppsala University, where I have worked with the Design of Materials for Energy Applications. During my Ph.D., I have published articles in high-profile Journals such as Advanced Materials (IF: 30.849), Small (IF: 13.281 ), The Journal of Physical Chemistry Letters (IF: 6.213), and JPCC (IF: 4.536). Currently, I am a Postdoctoral Researcher in Artificial Intelligence at Linköping University, Sweden, and Member of the Advisory Board at the Junior Faculty Steering Committee (LiU). I have also Founded Aicavity (E-learning Platform), and Oxaala Technologies, an AI software company focused on consulting with Offices in Brazil and Sweden. Additionally, I am Head Manager at Stockholm AI, a non-profit organization with AI enthusiasts, professionals, and researchers aiming to establish Stockholm as a global player in AI. I’m also a reviewer for the Journal Vacuum IF: 3.67 (Elsevier) covering Materials Science & Machine Learning.
Education & Academia
PostDoc in Artificial Intelligence
Development of Artificial Intelligence models.
Steering Committee Advisory Board Member
Licenciate and Ph.D. in Physics
My Thesis: Modelling and Simulation of Electro-catalysts for Green Energy: From Solvated Complexes to Solid-Liquid Interfaces
• Teaching Experience: Mekanik Baskurs (2 years)
• Administrative Experience: Responsible for the Materials Theory Division Seminars.
• Resources: My Ph.D. was financed by Uppsala University, and the projects/travels were mixed - Supervisor Dr. Moyses Araujo, and Co-Supervisors Dr. Olle Björnholm, and Dr. Barbara Brena (i.e. STINT-CAPES Brazil-Sweden, Knut and Alice Wallenberg Foundation, and Vetenskapsrådet).
• Computational Resources: KTH Beskow HPC, Tetralith SNIC, and Kebnekaise HPC. These projects involved approximately 300,000 cores*hours/month. I have written the project that was approved by the HPC Facility SNIC (JPCL 2019 Published).
• Experience: I have led the computational simulations of six projects using high-performance computing facilities to design materials for green energy.
•Programming Languages used: C/C++, Python, Matlab, and Shell-script.
• Metrics: 6 high-profile publications in international Journals.
• My Presentations:
Workshop: Trieste - Uppsala Meeting on Photoelectron Spectroscopy, Uppsala University, Sweden 2018. (Seminar)
CoTXS mini-workshop, Uppsala University, Sweden 2018. (Seminar)
SUNCAT Summer Institute, Fundamentals and Applications of Heterogeneous Catalysis, Stanford University, California, USA 2017 - Poster
European Congress and Exhibition on Advanced Materials and Processes EUROMAT2019, Stockholm, Sweden 2019 (Seminar) - Photo
College on Multiscale Computational Modeling of Materials for Energy Applications - ICTP, Trieste, Italy 2016 - Photo
BSc and MSc in Physics
• Lecturer in Theoretical and Experimental Physics I and II (UFBA/Devry University)
• Lecturer in Mechanics of Materials I and II (Devry University)
• Assistant Lecturer in Solid State Physics I (UFBA)
• MSc in Physics (UFBA) - Ranked 1st - Group: LaPO (Laboratory of Optical Properties) - Field: Solid State Physics and Computational Simulations. Group Leader: Prof. Antonio Ferreira. Supervisor: Prof. Denis Gilbert Francis David
• BSc in Mining and Petroleum Engineering (UFBA) (Completed 16 Courses in 1 year before moving to Physics) - Ranked 1st
• Marketing Director at Crystal Mining and Petroleum - Junior STARTUP (UFBA)
• BSc in Geophysics (UFBA) (Completed 2 years before moving to Sweden) - Ranked 1st
• Ph.D. in Geophysics (UFBA) (Completed 1 year before moving to Sweden) - Ranked 1st
• Junior and Master Scientific Research Scholarship (2,5 years BSc and 2 years of MSc - National Council for Scientific and Technological Development, CNPq, Brazil)
1 - Project: Simulation of CO2 and water injection in front reservoir using Apollonian Networks as a porous-media to improve oil recovery mechanisms. (Ansys Fluent) - Group: Statistical Physics and Complex Systems, Supervisor: Prof. Roberto Andrade (UFBA)
2 -Project: Study of Magnetic Properties in Paramagnetic Aggregates via First Principles Calculations - Supervisor: Prof. Roberto Rivelino (UFBA)
Certificates before 2015:
Conferences and Events
- School on Electronic Structure and Quantum Transport Methods, UNESP, ICTP, SP, Brazil, 2014.
- V Winter School of Physics - UFBA, BA, Brazil, 2012.
- Certificate in Thin Films Applied to Solar Energy.
- Certificate in Study of Semiconductor Materials via DFT.
- Certificate in Electron and ion accelerators.
- Certificate in Molecular Orbital theory.
- IV Winter School of Physics - UFBA, BA, Brazil, 2011.
- Certificate in Topics in Field Theory.
- Certificate in Solar energy.
- Physics Summer School at Aeronautics Institute of Technology (ITA). Seminar: Study of Magnetic Properties in Paramagnetic Aggregates via Calculation of First Principles, ITA, SP, Brazil, 2011.
- Winter School in Physics Applied to Medicine and Biology - USP, SP. 2011.
- AMS/SBF First Joint Meeting - IMPA, RJ. 2008.
- International Symposium of Dynamical Systems -. IMPA, RJ. 2006.
- V Workshop on Molecular Physics and Spectroscopy - UFBA, BA. 2006.
- XIV School of Differential Geometry - Honors to Shiing-Shen Chern (IMPA) - UFBA, BA. 2006.
Industry & Non-Profit
Founder & CEO
Non-profit Artificial Intelligence community in Stockholm.
Professional Network & Training.
Research Engineer (Summer Intern)
Deep Neural Network Architectures and Markov Decision Processes for tracking multiple objects.
Recent Research Projects
AI for Materials Discovery
- Development of DL and Graph Neural Network Architectures for high-throughput design and discovery of new molecular and 2D-based Materials with applications in Energy devices & Catalysis.
- Graph-Nets, Machine Learning and DL-based approaches for tracking possible switching mechanisms of molecular complexes with applications in Neuromorphic Computing.
- Discovering new Organo-Metallic coordinated Complexes for efficient Artificial Photosynthesis. Here I use GANs and SMILES to generate a Database. The DL Architectures are being used to track the pKA and reaction-based descriptors that can contribute to the prediction of efficient systems for Hydrogen Production. (Inspired by my Ph.D. Thesis)
AI for Real Simulations
- Applications of Graph Nets and Graph Neural Networks to simulate physical behavior of real particles and systems.
AI & Decision-Making
- Historical & Real-time Geophysical and Economic data-driven analytics. (Partners: CPRM, Petrobras and GETA UFBA - Brazil)
- AI for Data-driven Decision-Making, Operations Research, Options Pricing, Portifolio Management, Arbitrage, Valuation, and Investments.
- Framework for Detection and Tracking using DL Architectures, DRL and Markov Decision Processes.
- Development of AI-based frameworks for detection of anomalies in ECG and MRI images.
Tetralith, SNIC, Sweden
- Tetralith consists of 1908 compute nodes each with two Intel Xeon Gold 6130 cpus with 16 CPU cores each, giving a total of 61056 CPU cores.
- The performance of the complete system is around 3 Pflop/s (LINPACK Rmax).
Kebnekaise Cluster, Sweden
- 432 nodes with Intel Broadwell E5-2690v4, 2x14 cores, 128 GB/node
- 20 nodes, 3TB/node, Intel Broadwell E7-8860v4, 4x18 cores
- 32 nodes with 2x NVIDIA K80; 4 nodes with 4x NVIDIA K80; 128 GB/node
- 36 nodes, Intel Knights Landing; 7250 SKU (68 cores, 1.4 Ghz/1.2 Ghz/AVX), 192 GB/node.
- 52 nodes with Intel Skylake Gold 6132 2x14 cores, 196 GB/node
- 10 GPU Volta nodes with Intel Skylake Gold 6132 2x14 cores, 2xNvidia V100 (2x5120 CUDA, 2x640 Tensor), 192 GB/node
- InfiniBand EDR/FDR interconnect
Rackham Cluster, Uppsala University
- Rackham comprises 9720 cores in the form of 486 nodes with two 10-core Intel Xeon V4 CPU's each. There are "fat" nodes,4 with 1TB memory and 32 with 256 GB memory, with the rest having 128 GB.
- Rackham's storage system (named Crex) uses the Lustre file system and provides 6.6PB of storage.
- The interconnect is Infiniband FDR which supports a theoretical bandwidth of 56Gb/s and a latency of 0.7 microseconds.
AI for Trading Nanodegree
Deep Learning Specialization
Natural Language Processing Specialization
Natural Language Processing with Attention Models
Natural Language Processing with Classification and Vector Spaces
Natural Language Processing with Probabilistic Models
Neural Networks and Deep Learning
Structuring Machine Learning Projects
Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
Convolutional Neural Networks
Build Basic Generative Adversarial Networks (GANs)
AI for Medical Diagnosis
Natural Language Processing with Sequence Models
Introduction to Software Product Management
Software Processes and Agile Practices
Fundamentals of Reinforcement Learning
Sample-based Learning Methods
Introduction to Data Science
Fundamentals and Applications of Heterogeneous Catalysis
Poster: Nanodevices as micro-reactors for studying electrocatalytic hydrogen evolution reaction