Contact
ABOUT ME
Senior Machine Learning Engineer, Ph.D.
talkto@rubenszimbres.phd
I am a full-stack ML Engineer, certified Google Cloud Data Engineer and Google Developer Expert
(GDE) in Machine Learning (NLP) and Google Cloud Platform (Security), and AWS Certified Cloud
Practitioner, and AWS Certified Machine Learning - Specialty. I am CompTIA Security+ certified.
I have successfully led both Machine Learning and DevOps teams. mentoring staff, guiding
project efforts, and consistently meeting troublesome deadlines with efficiency. My approach
involves anticipating potential challenges and proactively implementing solutions to ensure
smooth project execution. I love dynamic environments, and my experience has equipped me
with the strategic mindset and leadership qualities necessary to drive innovation and foster a
collaborative team culture, proven by two patents of mine.
SKILLS
Python, Machine Learning, Deep Learning, Keras, Tensorflow, PyTorch, NLP, Generative AI, LLM,
LoRA, PEFT, LangChain, RAG, Knowledge Graphs, Time Series, XGBoost, SQL, Kubeflow, AWS,
Google Cloud, Security, Recommender Systems, Linux, LLMOps, MLOps, VertexAI, Sage Maker,
Bedrock, Cloud Architecture, ML Pipelines and Cybersecurity.
PDF Version
Resume Rubens Zimbres_.pdfPROFESSIONAL EXPERIENCE
Google for Startups (May 2024 - Present) - Mentor of Google for Startups Accelerator Brasil
Intellimetri (Apr 2019 - Jun 2023) - Senior Data Scientist / Machine Learning Engineer
Intellimetri is a spin-off of Vecto Mobile and Google Cloud Partner. We implemented end-to-end Google Cloud solutions using Vertex AI, Matching Engine and also customized Machine Learning and Deep Learning Algorithms. My focus was on Natural Language Processing and Recommendation Systems following best practices for cloud planning and deployment in ML solutions. At Intellimetri I was responsible for ML algorithms, pipelines, MLOps, end-to-end architecture design and costs of Google Cloud infrastructure. During this time I interacted with Google’s PSO specialists to discuss best practices for the Google Cloud environment.
Some projects I delivered:
A Two-stage recommender system of a Virtual Career Center solution, deployed in Vertex AI endpoints via Kubeflow pipelines, trained and deployed daily. The solution returns Top 10 job vacancies according to the semantic similarity with the profile of the user (candidate). This solution was delivered for a Colombia Municipality.
A highly scalable Recommender Model (retrieval) for a government agency in Mexico, deployed in Vertex AI Matching Engine.
A chatbot in Dialogflow where the users were able to search the content of brazilian laws, using a retrieval model based on similarity of embeddings (Two Towers) and deployed in Cloud Run container via Container Registry.
A computer vision solution for supermarkets that was able to identify and count milk cans at the cashier, in order to avoid fraud. The object detection was made with Tensorflow.
Revenue prediction using Temporal Convolutional Networks in Tensorflow for a hotel chain with 130 units. The average precision achieved for the next 3 months forecast was 91.5%.
Proof of concept of retrieval models for a career solution for governments of Middle East, deployed in HuggingFace Spaces using Gradio interface.
Call center solution: Speech analytics and automation of voice call monitoring with Machine Learning.
Algorithmic trading: a solution was deployed with Compute Engine and cron jobs for swing trading, considering financial indicators as MACD, Moving Averages, Fibonacci and RSI for BOVESPA stock exchange.
We generated two patents in Natural Language Processing at INPI (BR512020002649-4 and BR512023000848-6) in 2020 and in 2023. I was also in charge of Research and Development of NLP solutions, including retrieval models and Transformers in PyTorch and Tensorflow. During this period, I mentored a co-worker, an undergraduate student in Applied Mathematics at University of São Paulo, in his final course work entitled "Natural Language Processing in the automatic analysis of Resumes, Videos and Virtual Interviews", using Google Cloud Technologies, where the student scored 10/10 according to the evaluation board.
Google Developers (Oct 2022 - Current) - Google Developer Expert in Machine Learning and Google Cloud Innovator
I contribute to the Google Developers Experts program by developing innovative solutions combining Google Cloud APIs and services to solve business needs using Machine Learning and Deep Learning, implementing and deploying in Google Cloud. Then, I write articles regarding my experience in Medium blogs, like Google Cloud Brazil and Google Developers Experts. As Google Cloud Innovator, I test Google Cloud new technologies.
Google Developers (May 2022 - Current) - Google Developer Expert in Google Cloud Platform
Vecto Mobile (Oct 2017 - Apr 2019) - Data Scientist
I was responsible for Research and Development of Machine Learning and Deep Learning algorithms, new products and solutions involving speech recognition and prediction models. Due to our innovative efforts in the Natural Language Processing field, specially for call centers, we became part of Google Cloud for Startups, Spark package. Six months later, we were upgraded to the Surge package. Then, we became Google Cloud Partners.
We deployed a virtual career solution in 6 countries of Latin America, that uses face recognition, sentiment analysis, speech analytics, and NLP. As a result, recruiters got up to 73% of average screening time reduction in selection processes, up to 78% of vacancies replacement time reduction and more than 85% of accuracy in choosing candidates. As of December 2022, the platform was used by more than 365,000 people.
Other projects I delivered:
Predictive Maintenance: by analyzing telemetry of Point-of-Sales (POS)
Scientific Coffee Roasting: IoT project where temperature and oxygenation of roaster were manipulated to generate a coffee with better taste, sweetness and scent
Fraud detection using voice, Speech-to-text and word embeddings for a Fraud Consulting Firm. The algorithm used was Gradient Boosting.
An end-to-end streaming IoT project to collect and analyze telemetry in real-time: made in AWS using IoT Core, Amazon Kinesis, Firehose, DynamoDB.
Google - UOL Diveo (Mar 2019) - Lecturer
I presented our successful business case at Google - UOL Diveo Conference for managers and directors of big retailers in Brazil. The business case was about a solution using Natural Language Processing, Speech Recognition and Deep Learning deployed in a Google Cloud infrastructure.
Goiás University (Feb 2018) - Instructor
Instructor at Deep Learning Brazil Summer School, where I lectured about GANs (Generative Adversarial Networks) and presented a workshop of practical implementation in Keras.
Doux Dermatology (Oct 2010 - Jul 2013) - Business Intelligence Analyst
I doubled the business profit in the first year using strategies of digital marketing, market research and customer focus. After 2.5 years net profit increased 239%, financial leverage decreased 36%, stock turnover increased 15%, ROI increased 206%, ROE 95% and EBITDA 98.5%.
DR3 Consulting (Feb 2005 - Dec 2009) - Consultant
My consulting firm specialized in strategic management and analysis of market research. The goal was to identify businesses' needs and customer perception of quality, to guide strategic approaches in market segments. One of our clients doubled the dental clinic facilities area after 2 years.
Brazilian Air Force (1995 - 1997) - 1st Lieutenant Dentist of the Air Force
I was promoted to the Head of the Health Sector and I was responsible for military training, facilities security and also meetings with high ranking officials to define logistics strategy for the health sector. I was a volunteer on a mission to rescue Brazilian tourists on the island of St Maarten (Caribbean), victims of Hurricane Luís.
EDUCATION
Brown University (2006)
Wolfram Research Summer School. I studied the effects of neighborhood in the cellular automaton model of my Doctorate thesis using Wolfram Mathematica. Results were published at Electronic Notes in Theoretical Computer Science - Elsevier and presented at a brazilian national congress of Business Administration.
Mackenzie Presbyterian University (2006 - 2009) - Dual Doctorate, Business Administration and Electric Engineering
I continued my Master's degree research using Agent-Based Modeling in Management to search for a cellular automata rule that could simulate human interactions in business services. Descriptive statistics, linear regression and Factor Analysis were used to analyze quantitative data. The results allow businesses to simulate future behaviors of clients and social network influencers so that firms can be more efficient in their marketing and management efforts. I was awarded a full 4-year research scholarship (merit) and additional training was sponsored by Wolfram Research, Illinois. In 2009, a patent was generated at INPI (RS 10950-1) and an article was published at Elsevier - Electronic Notes in Theoretical Computer Science, 2009. [ISBN: 1571-0661] and another article was accepted in the Academy of Management Annual Meeting, Montreal, Canada, 2010.
Mackenzie Presbyterian University (2004 - 2005) - Dual Master, Stricto Sensu, Business Administration and Electric Engineering
I started an innovative research using cellular automata to simulate human rationality and interactions and simulating them as social networks. The research was applied to dentists and a quantitative market research was conducted. Awarded full 2-year research scholarship, by merit.
The dissertation was entitled "Dynamics of formation and evolution of business networks in Dentistry". Articles were published at EnANPAD (Encontro Nacional dos Programas de Pós-Graduação em Administração de Empresas - Brazil), and in the 10th International Conference on Enterprise Information Systems (ICEIS 2008, held in Barcelona, Spain, June 12-16), INSTICC Press: Setúbal-Portugal, Vol. III: Artificial Intelligence and Decision Support Systems, pp. 333-339, 2008. [ISBN: 978-989-8111-37-1]
São Paulo University of Economics (2005)
I attended classes during 6 months, where concepts of Economics, Complexity, Game Theory and computational modeling were approached. I developed a model of human interactions based on The Standing Ovation Problem, of Miller and Page.
São Paulo University of Dentistry (1992)
PROFESSIONAL TRAINING
Machine Learning at Coursera and AI Overlords:
Supervised and Unsupervised Machine Learning, Deep Learning, Natural Language Processing, Recommendation Systems, Time Series Forecasting, Transfer Learning, Graph Neural Networks, CNNs, LSTMs, GRUs, GANs, Transformers.
Google Cloud Training at Qwiklabs: Cloud Architect track, Cloud Security Engineer track, Machine Learning Engineer track, Data Engineer track, Tensorflow, Streaming Analytics, Recommendation Systems
Agent-Based Modeling at Santa Fe Institute: Modeling with NetLogo, Python and Arduino
Cybersecurity at TCM Security, TryHackMe and Pluralsight:
Web Application Security and Pentesting, Malware Analysis, Reverse Engineering, Offensive Pentesting, Ethical Hacking, OSINT, Privilege Escalation (Linux and Windows), C Programming, Buffer Overflow and Active Directory Reconnaissance, Scanning, Enumeration and Exploitation.
Credentials at: https://www.linkedin.com/in/rubens-zimbres/details/certifications/
PUBLICATIONS
Cellular automata model of social networks (Version 1.0.0). CoMSES Computational Model Library, 2022. Available at:
https://www.comses.net/codebases/6f05df1d-0e51-403c-a5e8-45a41c99875a/releases/1.0.0/
Dynamics of quality perception in a social network: A cellular automaton based model in aesthetics services. Electronic Notes in Theoretical Computer Science, Elsevier [ISBN: 1571-0661] · Oct 1, 2009. Available at:
https://www.sciencedirect.com/science/article/pii/S1571066109003740
Cellular automata based modeling of the formation and evolution of social networks: A case in Dentistry. Artificial Intelligence and Decision Support Systems, pp. 333-339, 2008. [ISBN: 978-989-8111-37-1] · Jun 16, 2008.
PATENTS
Solution for search and compatibility of resumes of candidates and job vacancies using Artificial Intelligence. BR512023000848-6 · Issued Apr, 4, 2023.
Natural Language Processing algorithm for analysis of Contact Center calls. BR512020002649-4 Issued Dec 1, 2020.
Cellular Automata Code to simulate artificial societies and social networks in C++ and Mathematica. BR RS 10950-1 · Issued Feb 1, 2011.
LANGUAGES
English - Fluent
Portuguese - Native
Spanish - Advanced