//assets.cakeresume.com/assets/contentarea/contentarea-2d64eca8dae2a1ffb31e0fb1df887bbad072c7c19734d7d30fa14884bbcc299a.css

A search engine contains more than 2 thousand News from BBC, they were pre-Analyzed using an AI (from Google). It will search and show the news related or link to the keywords given in the search bar

AI - Search Engine 

Data scientist | Machine Learning | AI

Marvin G.

End to End Machine Learning Project + CI\Cd Implementation.

End to End Machine Learning

1. Utilize the AI model to Create new lyrics.
2. Developed a Deep learning model that mimics Canserbero's Writing Style.
3. Generate New lyrics by completing given sentences.

Generating New Lyrics 

Mobile App to help user to manage efficiently their time on social media.

Tech Vise

Docker for Machine Learning and Data Science

Udemy

Supervised Machine Learning

Data Analyst Foundation

Google

Unsupervised Machine Learning 

Google Data Analytics Professional Certificate

An MBA-MSIM-Certified Notebook Expert in Kaggle equipped with an in-depth understanding of the modern business environment and deployment of cutting-edge models under CI/CD methodology.

*** Dear recruiter do not be shy and say "hi" ***

National Taiwan University of Science and Technology

Universidad Nacional de Ingenieria Nicaragua

緯創資通股份有限公司

Machine Learning Engineer 

• Design and develop Intelligent chatbots, integrated with a center knowledge information system, and Open-AI.
 • Collaborate with cross-functional teams to measure NLP model performance. 
 • Promote the use of data science and machine learning techniques in a Site-Program, delivering 6 hands-on machine learning modules for nontechnical users.


台灣台北市

- Automatize 15+ systems by building and managing data pipelines that
crawl websites, NoSql-databases, and APIs for data collection.

- Debug and apply transfer-learning to existing machine/deep learning
models from pre-trained embedding: Glove, BERT, fastText, Universal
Sentences Encoder for text similarity & classification.

- Develop a semi-automated machine learning library built-in Talib &
Pandas libraries for exploratory analysis, forecasting, and stock
clustering.

Taiwán

Upwork as a Freelancer

Data Scientist

-Design and Build an Internal Search Engine to query Testing quality processes and procedures.
- Build a GUI interface to collect, process, and store information into a database, JSON files for (Search Engine), and Google Suit.
- Build and Troubleshoot a module to transform unstructured data into various formats Data includes python scripts for Office 365 and Google Docs transformation.

台灣新竹市新竹

Logitech 

Quality Systems Development

- Researched and prototyped: supervised, unsupervised algorithms and
state-of-the-art deep learning models: LSTM, RNN, BiLSTM for text
summarization, text generation, and classification.
- Partnered with NSTUT lab to enhance and tune NLP models for
Sentimental Analysis, Named entity Recognition, Keyword Extraction,
and Topic Modeling.
- Coached 4 members-data team through a short-term project to
construct and deploy a machine learning CRM data analytics system
reaching 90% in classifying customer purchases intention.

國立台灣科技大學 National Taiwan University of Science and Technology

Data Scientist Researcher

- Worked closely with CTO to set up the Graph-reduction clustering
software's modules for Spanish & English, achieving keyword extraction
in both languages.
- Designed and maintained internal text-preprocessing modules with built-in
NLTK, Spacy, and Keras libraries for Named-entity recognition,
information extraction, and text summarization.
- Built an internal compiler for python-Neo4J for graph visualization.

Taipéi, Taiwán

LnData  麟數據科技股份有限公司

Machine Learning Engineer

- Managed changes in project scope, key milestones, and performance indicators, involving a team composed of 3 software engineers, and 2
designers, for App - software development.
- Led translation of technical specifications, and communication with stakeholders in regards to detailed product requirements, features
concepts, and customer needs.
- Tracked risk management plan, feature development, bug troubleshooting, and delivered beta version within 6 months.

Share