André Diler Email & Phone Number

Montreal, Quebec, Canada

Establish the foundational framework for a user‑friendly youth book search engine, enhancing accessibility for educators and librarians to discover ideal educational content for children.• Consolidate data into RDF with Python and SPARQL queries in GraphDB to create a versatile knowledge database, aiming to further enable nuanced queries such as finding well‑rated books for 6‑year‑olds interested in whales.• Conduct code reviews and perform data integrity checks on team members’ deliverables.• Conduct a literature review to align book themes from multiple websites (thesauri matching), oversee the creation of a manually annotated ground truth dataset by reviewers following my instructions, and evaluate various techniques for comparison.

Successfully debug, optimize and modernize mission critical tools used in the trading of exotic macro economic products like multi‑legged interest rate swaps.• Own maintenance of a trading application used by hundreds around the globe by prioritizing issues, proactively fixing bugs and adding features in Scala & Java using a software infrastructure developed internally (Optimus, which is open source)• Increase performance of a feature by 9000% (from 6 minutes execution to 4 seconds)• Reduce time complexity of a feature from O(N) to O(1) resulting in hours of daily time savings for users and computation.• Provide technical expertise to hiring managers by conducting entry‑level technical interviews and assessing candidate qualifications.• Implement best practices for code reviews, testing, and deployment as a key contributor to a state‑of‑the‑art software development lifecycle

• Assist a biomechanical student group in building a classification model to determine different movements from ECG signals in order to build an exoskeleton by exploring and dispatching the most promising approaches in the literature into small teams• Lead a team that developed a new ECG signal processing system using a custom transformer• Created an adversarial images dataset fooling most computer vision models for the most difficult challenge of the CodeML 2020 hackathon

Canada

• Manage external relations with student committees and companies• Raised 4000$ for the second edition of the CodeML hackaton (CodeML 2020)

Montreal

Design and build a cutting‑edge service that ensure secure storage of bank access tokens in a serverless architecture, giving users the convenience of accessing all their bank accounts in one place without the need for re‑authentication.• Implement a streamlined CICD pipeline with BitBucket and AWS, enabling the swift and efficient building, testing, and deployment of our code in stacks of Lambdas, Databases, and API gateways.• Transform the dynamic IP address of Lambdas to a static ElasticIP using a NAT gateway, providing an added layer of security and enabling our service to be whitelisted by banks.

Montreal, Quebec, Canada

Develop a tool that significantly increased efficiency by reducing downtime and saving dozens of man hours per week through the centralization of machine upgrade issues and solutions for hundreds of users using SQL Alchemy and Python in Linux.

Remote

• Explore different approaches to increase the performance of algotrading algorithms with RNNs in Python, SQL and PyTorch• Reduce RAM consumption on Google Colab by lazy loading a database of daily indicators of 5000 stocks over 20 years

Montreal, Canada Area

• Preprocess and scale a complex dataset (DeepLesion)• Train machine learning models (manifold learning) to predict tumor response to chemo• Share code with team members• Create figures to visualize my results and methods

Montreal, Canada Area

Streamline the work of radiologists by automating the classification of tumors in liver cancer using CNNs in Python & Keras, which resulted in increased efficiency and accuracy as a member of a multi‑disciplinary team including doctors, medical students, data scientists, and PhD students.• Develop a software prototype to significantly improve the visualization of tumor evolution over time for radiologists using Python and Javascript, resulting in better diagnosis and treatment decisions.• Publish results (0.96 ACC) in a scientific publication as the first author, demonstrating the effectiveness of the developed model

Bachelor'S Degree, Computer Software Engineering

Exchange Semester, Computer Engineering

Exchange Semester