Álvaro Peris
Principal Engineer at LendingClub
Valencia, Valencian Community, Spain
Summary
👤
Senior🎓
Top SchoolÁlvaro Peris is a Principal Engineer and machine learning specialist with 11 years of experience building production-grade ML systems and end-to-end products for finance and consumer startups. He holds a PhD in machine learning focused on deep learning for multimodal sequence-to-sequence modeling and HCI adaptation, and his work blends research rigor with pragmatic engineering. At Cushion he built a high-throughput email understanding pipeline (processing up to 300K emails/day) that combined fine-tuned, quantized LLMs with robust reconciliation and SKU-level extraction, achieving over 95% automation. His open-source contributions to neural machine translation tooling (nmt-keras) reflect deep expertise in sequence models, attention mechanisms and training optimizations. Comfortable wearing many hats—from software engineer to data engineer—he excels at owning complex projects from research prototype to reliable production. Based in Valencia, Spain, he pairs startup agility with academic depth to solve data-rich, messy real-world problems.
11 years of coding experience
9 years of employment as a software developer
Engineer's degree Computer Science, Engineer's degree Computer Science at Hochschule für Technik und Wirtschaft BerlinGerman Studies, German Studies at Centro Alemán ValenciaUniversitat Politècnica de València
Spanish, English, Catalan, German
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (7)
keras10
machine-learning10
lstm10
deep-learning10
nmt10
python10
tensorflow9
Programming languages (3)
Github contributions (5)
lvapeab/nmt-keras
Nov 2016 - Jul 2021
Neural Machine Translation with Keras
Role in this project:
Back-end Developer
Back-end DeveloperContributions:12 releases, 968 commits, 17 PRs in 4 years 9 months
Contributions summary:Álvaro's commits primarily involve modifying and refactoring the core files related to neural machine translation with Keras, including significant changes to the main training script, the translation model, and associated utility files. They have introduced features related to the selection and implementation of different RNN units, including LSTM and GRU, as well as options related to attention mechanisms like adding support for transformer architectures. Their work also touched upon managing and utilizing the models' training parameters such as gradient accumulation and more granular control of the training parameters.
kerasneural-machine-translationdeep-learningsequence-to-sequencetheano
lvapeab/coco-caption
May 2016 - May 2019
Contributions:1 release, 42 commits, 35 pushes in 3 years
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.