Rubert Martín Pardo Email & Phone Number

Montreal, Qc, Ca

Montreal, Quebec, Canada

Montreal, Quebec, Canada

As a Sr ML Engineer, my duties extended to the the operational side of deploying ML models.Key responsibilities:- Design and test vector search service using bi-encoders and cross-encoders for retrieval and reranking phases. Improves job matching using semantic search.- Develop APIs for our Machine Learning models to serve requests coming from both Development and Production environments.- Manage the infrastructure that automatically delivers Machine Learning inference endpoints to multiple AWS accounts, under different requirements of scalability and latency. - Create libraries in multiple languages (Golang, Typescript, Python) that provide user-friendly interface to our ML models.- Automatize the data cleaning, and extraction from raw sources for our NLP-base models to predict and classify Named Entities in the online recruitment industry.

Montreal, Quebec, Canada

• Natural Language Processing Specialist. Machine Learning Engineer. § Developed Named Entity Recognition (NER) models that help detect key data such as 'occupation' and 'company name' from job ads and job titles for better employer -- employee matching.§ Consolidated the NER model across the company obtaining 95% accuracy and 3X speed relative to old rule-based model.§ Segmented and classified job ads via Hugging face NLP models into commonly seen blocks such as 'offer' and 'company description' to better classify and process online job postings.§. Led the testing, interviewing and hiring of four Machine Learning interns and graduate students for delivering on the support of numerous initiatives such as 'User intent prediction' and 'Multilingual NER'.** Skills** Named Entity Recognition models: rule-based models, transformer models (BERT family). Spacy and Huggingface frameworks. Development, evaluation, and deployment in AWS Sagemaker. Queries and ETL with AWS Athena and Glue. Parallelization.Document semantic segmentation and classification. Multiple languages standardized datasets. Zero-shot learning for languages with no/few-label datasets.

Montreal, Quebec, Canada

Thesis: Fluid dynamics of liquid surface instabilities.- Developed an analytical model that predicts jetting onset on impulsive flows around submerged structures.- Conducted novel experiments to validate the analytical model. Obtained prediction accuracies above 95%.- Developed machine learning and edge detection algorithms for analysis and classification of surface perturbations and detection of jetting onset.- Dissemination of key results in conferences and scientific publications.

Montréal, Canada

FACC 400 Engineering ethics- Organiced and delivered tutorials, evaluations and surveys for a class of 300+ students each semester.- Case studies on Nuclear Engineering, Cryptocurrency, Artificial Inteligence, among others.

Montreal, Quebec, Canada

- Advanced the analysis of data and prediction of trends for a dataset for the organisation Maisson des Femmes.- Collaborated with a multidisciplinary team to provide solutions and decision-making for NGOs .

Montreal, Canada Area

- Organized and taught at the McGill Physics Olympiad Program (https://physicsmatters.physics.mcgill.ca/mpop/)- Trained highly motivated high-school CEGEP students to particiapte in Physics Olympiads at the national and international level: Sir Isaac Newton Physics Olympiad, Canadian Physics Olympiad, International Physics Olympiad.

Burnaby, British Columbia, Canada

- Development of diagnostic tools to detect surface instabilities in the interior of a fusion reactor.- Development of experimental and numerical methods to describe liquid instabilities when pushed through a mesh.- Analysis of Taylor-Couette flow and torque scaling in a liquid rotating apparatus.

Argentina

- Contributed to the development of a real-time nuclear reactor simulator.- Utilized parallel-programming techniques to optimise the temporal response of the simulator.- Extended the simulator to multigroup energetics of the neutronics of the reactor.

Río Negro Province, Argentina

- Coupled a thermal-hydraulic code (RELAP) to a neutronic code (PUMITA).- Quantified the thermal coefficient of Saudi Arabia nuclear reactor designed by INVAP.

Havana

Mathematics, Experimental Physics

Doctor Of Philosophy - Phd, Engineering

Thesis: Prediction and classification of fluid instabilities and jetting onset. Relevant courses (grade): - Fundamentals of Deep Learning.

Master'S Degree, Nuclear Engineering

Thesis: Improvement of a nuclear reactor real-time simulator for training of operators. Relevant courses: - Computational Fluid dynamics..

Bachelor'S Degree, Nuclear Engineering

Thesis: Neutronic-thermohydraulic coupling of nuclear reactor modelling codes. Relevant courses: - Numerical methods in engineering..