Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

🤖 I'm an applied mathematician turned deep learning researcher turned AI engineer passionate about building software tools powered by AI. 📐 I started working on artificial intelligence in 2017 when I went to study at McGill University and started doing research on optimisation algorithms for deep learning. 📽 I have experience with both deep learning based computer vision and LLMs, where I have done everything from data collection, to model training, fine-tuning and deployment. 🎙️ Recently, I started expanding out of ML and learned Ruby on Rails for backend and Typescript/React for frontend.…🤖 I'm an applied mathematician turned deep learning researcher turned AI engineer passionate about building software tools powered by AI. 📐 I started working on artificial intelligence in 2017 when I went to study at McGill University and started doing research on optimisation algorithms for deep learning. 📽 I have experience with both deep learning based computer vision and LLMs, where I have done everything from data collection, to model training, fine-tuning and deployment. 🎙️ Recently, I started expanding out of ML and learned Ruby on Rails for backend and Typescript/React for frontend.

Mariana Oliveira Prazeres

Full Stack Developer, Back-End Developer, Web Developer, +4





🤖 I'm an applied mathematician turned deep learning researcher turned AI engineer passionate about building software tools powered by AI. 📐 I started working on artificial intelligence in 2017 when I went to study at McGill University and started doing research on optimisation algorithms for deep learning. 📽 I have experience with both deep learning based computer vision and LLMs, where I have done everything from data collection, to model training, fine-tuning and deployment. 🎙️ Recently, I started expanding out of ML and learned Ruby on Rails for backend and Typescript/React for frontend.…🤖 I'm an applied mathematician turned deep learning researcher turned AI engineer passionate about building software tools powered by AI. 📐 I started working on artificial intelligence in 2017 when I went to study at McGill University and started doing research on optimisation algorithms for deep learning. 📽 I have experience with both deep learning based computer vision and LLMs, where I have done everything from data collection, to model training, fine-tuning and deployment. 🎙️ Recently, I started expanding out of ML and learned Ruby on Rails for backend and Typescript/React for frontend.

Available to hire

🤖 I’m an applied mathematician turned deep learning researcher turned AI engineer passionate about building software tools powered by AI.

📐 I started working on artificial intelligence in 2017 when I went to study at McGill University and started doing research on optimisation algorithms for deep learning.
📽 I have experience with both deep learning based computer vision and LLMs, where I have done everything from data collection, to model training, fine-tuning and deployment.
🎙️ Recently, I started expanding out of ML and learned Ruby on Rails for backend and Typescript/React for frontend.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Language

Portuguese

Fluent

English

Fluent

Danish

Advanced

Spanish; Castilian

Intermediate

French

Intermediate

Work Experience

Chief Technology Officer (CTO) at Meet-Ting

July 1, 2025 - Present

AI: Translate open-ended product requirements into formalized AI problems, design evaluation criteria, and iterate on model behavior through controlled experimentation. Fullstack, Backend, & Infrastructure: Design and implement scalable backend services in TypeScript (NestJS). Deploy and manage systems on GCP, integrate with HubSpot and Amplitude, and cover the full stack from system design to monitoring and alerting. Technical Leadership: Define the technical roadmap, make architectural decisions across AI and backend systems, and contribute production code alongside other engineers in a fast-paced research-to-product environment.

Career Break / Independent Projects at Independent Projects

March 1, 2025 - July 1, 2025

Authored a newsletter on AI evaluation and applied machine learning practices: https://theevalloop.substack.com/. Explored freelance opportunities and experimented with independent project development.

AI and Backend Founding Engineer at Lex

November 1, 2023 - February 1, 2025

Led all AI-related development across the platform, including data collection, model fine-tuning, and systematic performance evaluation. Hands-on with multiple LLM APIs, retrieval-augmented generation (RAG) pipelines, and experiment tracking with LangSmith. Backend & Fullstack Development: Designed and implemented backend services in Ruby on Rails, including search and indexing (Elasticsearch), user authentication (WorkOS), notifications (Knock.app), and shared folder systems to support collaboration. Technical Leadership: Contributed to architectural decisions and development planning, collaborating with cross-functional teams to deliver production-ready features and maintain system reliability.

Deep Learning Engineer at Uizard

November 1, 2021 - November 1, 2023

Computer Vision: Training CNNs both from scratch and by fine-tuning ImageNet pre-trained models on proprietary datasets for object recognition & image classification problems. Experience with both PyTorch and TensorFlow. NLP: Creating agents and tools using GPT and Langchain to solve problems in the design-domain. Model evaluation, results visualisation and full pipeline deployment using internal tooling.

Invited Researcher at Inductiva Research Labs

July 1, 2021 - September 1, 2021

Training physics inspired neural networks from scratch on CFD-generated data to solve problems in mechanical engineering.

Machine Learning Intern, Search Team at Gympass

February 1, 2021 - June 1, 2021

Created the first machine learning model for gym search going from data collection to ElasticSearch staging prototype.

Associate Researcher at Huawei

June 1, 2020 - December 1, 2020

Implemented a regularization penalty for ResNets that resulted in 90% of final accuracy being achieved after only 50 epochs on CIFAR10; Implemented multiplication replacement in convolutional layers that led to 2x speed-up on inference time without decrease in accuracy on CIFAR10 and ImageNet. Presented and published at conference. Patent under review.

PhD Researcher / Neural Networks at McGill University

September 1, 2017 - April 1, 2020

Conducted research on the mathematics of neural networks applied to computer vision problems; analyzed optimization methods, including stochastic gradient descent with Polyak’s learning rate, and analyzed convergence for strongly convex functions. Developed Python, PyTorch, deep learning, computer vision, and AI system evaluation skills; published in Journal of Scientific Computing; teaching assistant duties in linear algebra and geometry.

Teaching Assistant at McGill University

January 1, 2018 - April 1, 2018

Teaching Assistant duties in linear algebra and geometry.

Data Science Fellow at DS4A by C1

September 1, 2021 - October 1, 2021

Data Science Fellow.

Deep Learning Research Intern at Uizard

September 1, 2019 - February 1, 2020

Deep learning research intern focusing on models and practical DL deployment.

Visiting Student at McGill University

June 1, 2016 - August 1, 2016

Visiting Student.