Ignacio Tampe Palma


Data Scientist @ TELUS


Computer Science Engineer (Masters Equivalent)
B.Sc. Computer Science

You can find my resume here!

Areas of Interest

Deep Learning

Deep Neural Networks, CNNs, Autoencoders, Attention, applied for supervised and unsupervised tasks with Pytorch and Tensorflow with training on Sagemaker

Data Science

ETL Processing using Python and Airflow, Statistical Analysis, Time Series Forecasting, Recommendation Systems and KPI-based solutions

Machine Learning

Support Vector Machines, Decision Trees, Random Forest, Clustering techniques, Feature Engineering using Sklearn

Text Mining

Word embeddings, Transformer Architectures, self attention, topic modeling applied on clasification, exploratory and, generative tasks using Huggingface and Gensim


Experience

Data Scientist - AI Accelerator
TELUS
Jul. 2023 - PRESENT
Vancouver, Canada
Data & Applied Scientist II
Microsoft
Sep. 2021 - Feb. 2023
Vancouver, Canada
  • Design, model, and develop models to assists the insights provided by Clarity.
  • Successfully developed and deployed a feature that generates automatic insights digests for customers with recommendations and metrics, reaching over 300k users.
  • Handle Big Data (~30TB) queries on High Performance Compute Clusters
  • Perform data analysis reports using python and C# to internal stakeholders.
Data Scientist
Cornershop by Uber
Dec. 2020 - Sep. 2021
Santiago, Chile
  • Implemented a deep learning forecasting tool for time series analysis to predict demand and capacity of the operation in the groceries team.
  • Perform MLops to deploy and serve models, with continuous integration using Airflow, SageMaker and Tensorflow Serving.
  • Collaborated with other data scientists to design a user segmentation and recommendation system.
Visiting Researcher
Dalhousie University
Jan. 2020 - Dec. 2020
Halifax, Canada
  • Natural Language Processing & Deep Learning Researcher at the Faculty of Computer Science in Dalhousie University under the ELAP Program.
  • Research focused on Automatic Abstractive Unsupervised Summarization of online news discussions.
  • Work culminated in a paper accepted at intelliSys 2021

Publications

Neural Abstractive Unsupervised Summarization of Online News Discussions
[Paper][Preprint] [Code]

I. Tampe, M. Mendoza, and E. Milios, Neural abstractive unsupervised summarization of online news discussions. . In Intelligent Systems and Applications: Proceedings of the 2021 Intelligent Systems Conference (IntelliSys) Volume 2 (pp. 822-841). Springer International Publishing.


Education

Computer Science Engineer M.Sc Equivalent (1.5 Years Track)
Mar. 2019 - Nov. 2020
Universidad Técnica Federico Santa María
  • Relevant courses: Deep Learning, Text Mining, Business Intelligence, Pattern recognition in data mining, Image Processing Fundamentals, Advanced Databases.
  • GPA: 80/100 | 3.7 (A-)
  • Thesis: "Neural Abstractive Unsupervised Summarization of Online News Discussions"
    Graded 100/100 | 4.0 (A+)
BsC. Computer Science (4 Years Track)
Mar. 2015 - Dec. 2018
Universidad Técnica Federico Santa María
  • GPA: 80/100 | 3.7 (A-)
International Baccalaureate Diploma
Mar. 2013 - Dec. 2014
  • Final Grade: 35

Undergraduate Experience

Software Engineer Intern
European Southern Observatory (ESO)
Jan. 2019 - Feb. 2019
Paranal, Chile
  • Developed a software tool to handle technical time and repairs for the telescopes.
  • Achieved improved efficiency, consistency and scalability compared to the previous process.
Data Structures & Algorithms Teaching Assistant
John's Hopkins Center for the Talented Youth (CTY)
Jun. 2019 - Aug. 2019
Los Angeles, CA, USA
  • Taught eighth grade students about Data Structures and Algorithms.
  • They successfully applied this knowledge to solve computer science challenges from concept to code.
Undergraduate Teaching Assistant
Universidad Técnica Federico Santa María
2016 - 2020
Santiago, Chile
  • Deep Learning.
  • User Interface Design.
  • Non-Relational Databases.
  • Algorithms & Complexity.
  • Data Structures.
  • Computer Programming in Python.

Awards

Best Graduated Student Award 2021
2022
Chile

Every year the School of Engineers awards a Distinction to the graduate engineers of the last promotion of the Universities and Academies recognized by the School. This Distinction is granted to those engineers designated by the respective Universities, in recognition of their academic performance and their conditions of leadership and participation.

"Federico Santa María" Academic Distinction
2021
Chile

Highest Academic Distinction awarded to the best graduated student by GPA on the Computer Science Engineer class of 2021. The person who receives it stands out in achieving the profile of excellence, has recognized merits and consecration in their studies.

Emerging Leaders of America Program (ELAP)
2020
Canada

The Emerging Leaders in the Americas Program (ELAP) scholarships provide students from Latin America and the Caribbean with short-term exchange opportunities for study or research, in Canada, at the college, undergraduate and graduate levels. The grant consisted of 5500 USD.

Seeds for the Future
2019
China

Seeds for the Future is a 3 week immersive program based in China. Successful applicants spend an initial period at the Beijing Language and Cultural University. From there they are flown to Huawei’s HQ in Shenzhen where they learn about Huawei’s international business culture and receive technical training from some of the finest professionals in the industry.

Cool Projects

#ChileWokeUp
View Projector Demo (takes a bit to load)

Understanding the role of Social Media during the October 2019 Chilean protests, by analyzing the violence in posted images with deep learning and time series analysis. The linked demo shows some of our data projected in a vector space to find simmilarities based on ImageNet Embeddings.

Le Mi Know (When it runs)
View on Github
Downloads

A small open-source python library to get a notification when your function call starts, finishes, or it crashes during runtime, by only adding two additional lines of code. It currently needs an update :(

EmolBot
View on Twitter

A Twitter bot that posts fake news titles in Spanish, Works thanks to a deep learning encoder-decoder using a BERT architecture, came up as a side project using my thesis research, posts every three hours and generates an image accordingly.