Ismael T.

Data Scientist

660 dollar
Freelancer
11 years
Hermosillo, MEXICO

My experience

More

FreelanceJuly 2015 - Present

Developed personal projects such as: Recommender system for Santander product recommendation applying LSTM in Tensorflow. Developed advanced machine learning algorithm to make good decisions in online video games, improved the performance of players over 50%. Implemented a chat bot for twitch platform using Pircbot API. Assisted a Startup in the development of a predictive model of sale for food delivery.

Thermo Fisher ScientificDecember 2016 - Present

More

Tiempo DevelopmentAugust 2013 - August 2015

Implemented advanced machine learning algorithms in Java and Python to detect fraud and make decisions reducing the cost of previous solution in 100%.
Managed and coordinate the team in order to accomplish the deliveries on time handling CRISP-DM methodology in the life cycle of projects.
Identified the client concerns so as to propose innovative solutions and actionable business insights using their data through machine learning algorithms and statistical analysis.
Designed the evaluation metrics to achieve our objectives.
Generated charts to display the results in easy way to understand in non- technical levels using tableau, kibana or pandas.
Extracted, transformed and loaded data from Hadoop/Spark to machine learning algorithms in Linux environment.
Implemented Elastic Search to retrieve all documents in the company from one source.
More

Instituto Tecnológico de SonoraAugust 2012 - June 2013

Taught the subjects: data structure and numerical analysis in Java.
More

Hewlett-Packard LaboratoriesJune 2010 - July 2012

Research Assistant at HP Labs, Developed and researched algorithms for NLP projects using Word2Vec, Doc2Vec, Word Embeddings, Topic Modelling and the most popular libraries: NLP Stanford, Nltk, MiniPar parse, GATE. Collected Facebook and Twitter information talking about hp products using a web crawler. Devised a bag of words that contains the polarity of adjectives for assigning sentiment degrees to opinion words then applied those words in fuzzy logic algorithm so enhancing our results on sentiment analysis. Assigned rating to products using the opinions in order to develop a recommender system.

My stack

Big Data

Spark, Hadoop, Big Data

Analysis methods and tools

Agile Methodology

Other

Fuzzy Logic, Fraud Detection, Word2Vec, Statistics, Japanese, NLTK, Elastic Search, R, GATE, Business Analytics, Neuro-Linguistic Programming (NLP), Sentiment Analysis, NLP Stanford, Topic Modelling, English, Word Embeddings, Pircbot API, Business Intelligence (BI), MiniPar parse, Genetic Algorithms, Doc2Vec, Data Modelling, Microsoft Office, Data Structures, Python Programming

Databases

MySQL

Machine Learning

Neural networks, TensorFlow, Deep learning

Technologies

Machine Learning, Pandas

Others

Spanish, Strategic Planning, Twitter, Problem Solving, Data Science, Kibana, Business Analysis, Data analysis, Natural Language Processing (NLP), Business Strategy, Research, Numerical analysis, Project Management

Languages

Matlab, SQL, Python, Java

IT Infrastructure

Linux

My education and trainings

Computer Science, Artificial Intelligence - Universidad de Sonora2006 - 2010

- Benemérita Universidad Autónoma de Puebla2009 - 2010