Daniel Abib

Architected and built a full-stack AI video production platform using FastAPI, Temporal.io, SQLAlchemy, and Redis, with React-based UI for async video generation workflows. Deployed serverless ML inference on Modal.com and designed scalable GenAI microservices on GCP using Cloud Run and managed services.

Led training and optimization of Stable Diffusion XL (SDXL) using distributed training and LoRA fine-tuning, improving reproducibility with Weights & Biases and FiftyOne. Built dataset engineering pipelines with multimodal captioning and implemented scalable training/inference workflows on RunPod and Modal.com.

Fine-tuned Pulze’s Project Dream AI using Dreambooth, LoRA, and negative embeddings for architectural visualizations. Developed dataset curation pipelines and custom ComfyUI integration/tools using WebSocket for real-time monitoring and workflow automation.

Contributed to a Latin GRAMMY-nominated short form music video using AI-driven animation techniques. Animated characters across frames with Stable Diffusion using Dreambooth, LoRA, and Textual Inversion to maintain consistent attributes.

Produced digital video work including color grading and format conversion while optimizing editing workflows with FFMPEG and shell scripting. Ensured visual quality through strong composition, pacing, continuity, and consistency, later applying these skills to generative media workflows.

Studied for a Bachelor in Cinema and Audiovisual at Universidade Federal Fluminense from 2010 to 2015.

Activities and societies: Built foundations in linear algebra, multivariable calculus, and vector calculus; applied these concepts to machine learning algorithms and problems.

Completed the Mathematics for Machine Learning specialization on Coursera, strengthening linear algebra and multivariable/vector calculus and applying them to practical machine learning algorithms.

Activities and societies: Topics included supervised/unsupervised learning and best practices in machine learning; applied linear regression, logistic regression, and neural networks.

Completed Andrew Ng’s Machine Learning course on Coursera, covering supervised and unsupervised learning and applying algorithms such as linear regression, logistic regression, and neural networks.

Pursued a Bachelor in Information Systems at Universidade Federal do Estado do Rio de Janeiro (UNIRIO) from 2019 to 2020; the program was listed as unfinished.

Activities and societies: Used Python libraries (Pandas, NumPy, SciPy) for statistical analysis and data manipulation, including data visualization, probability, and inference.

Completed the Statistics with Python specialization on Coursera, using Python for data visualization, probability, and statistical inference with libraries including Pandas, NumPy, and SciPy.