Carlos Mocholí
Member Of Engineering at poolside
Spain
Summary
🤩
Rockstar🎓
Top SchoolCarlos Mocholí is a research-focused software engineer with 11 years of experience accelerating AI research and production workflows, currently a Member of Engineering at poolside after senior research roles at Lightning AI. He combines a Computer Science degree and an MSc from Edinburgh with deep practical expertise in algorithms, distributed systems, and machine learning, contributing to flagship projects like PyTorch and PyTorch Lightning to improve GPU acceleration, distributed training, and checkpointing. His open-source work spans dataset distillation and Lightning-Flash, where he extended data pipelines and prediction utilities—showing a blend of low-level performance fixes and higher-level ML tooling. Known for pragmatic bug fixes that unlock large-scale training efficiency, he excels at turning research ideas into robust, production-ready code. Based in Spain, he brings research rigor together with hands-on engineering to speed up AI experimentation and deployment.
11 years of coding experience
5 years of employment as a software developer
Master of Science - MS, Master of Science - MS at The University of EdinburghUniversitat Politècnica de València
German, Spanish, English
Github Skills (21)
pytorch10
distributed-training10
checkpoint10
python10
classification10
machine-learning10
pytorch-lightning10
model-checking10
checkpointing10
deeplearning-ai10
deep-learning10
gpu10
cuda10
debug9
debugging9
Programming languages (15)
Github contributions (5)
Lightning-AI/pytorch-lightning
Aug 2020 - Jan 2023
Pretrain, finetune ANY AI model of ANY size on multiple GPUs, TPUs with zero code changes.
Role in this project:
ML Engineer
ML EngineerContributions:17 releases, 9215 reviews, 961 commits in 2 years 5 months
Contributions summary:Carlos's commits focused on modifying the model checkpointing process, specifically adding and enabling the saving of checkpoints at the end of the training epoch, modifying metrics, and ensuring the last checkpoint is correctly saved. They also addressed a bug related to the automatic removal of old checkpoints. These contributions involved the modification of existing callback code and tests.
pythonheadachespytorch-modelsdata-sciencehandling
Lightning-Universe/lightning-flash
Jan 2021 - Jul 2022
Your PyTorch AI Factory - Flash enables you to easily configure and run complex AI recipes for over 15 tasks across 7 data domains
Role in this project:ML Engineer
ML EngineerContributions:150 reviews, 16 commits, 10 PRs in 1 year 5 months
Contributions summary:Carlos's contributions focused on improving and extending the functionality of the `lightning-flash` repository, a framework for rapidly developing AI recipes. Their work included fixing import errors, updating initialization files, and addressing minor issues. They also implemented a task.predict function to extend the usability of the framework and introduced a DataPipeline class to enhance data processing capabilities.
torch-geometricicevisionconfigurepytorch-lightningdomains
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.