Francisco Perez-sorrosal
Principal Research Engineer (Science Yahoo Knowledge Graph)
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolFrancisco Perez-sorrosal is a Principal Research Engineer based in San Francisco with 14 years of experience building scalable ML and distributed systems, currently leading entity reconciliation and knowledge-graph engineering at Yahoo. He combines deep research roots (PhD-level work on consistency and replication) with hands-on production ML: consolidating dozens of legacy models into a unified transformer-based reconciliation system that improved precision/recall substantially while cutting model size and operational complexity. His work spans taxonomy-driven multi-label classification, distillation for lightweight deployment, and cost-conscious cloud inference (including a Ray-based deployment PoC), reflecting a rare mix of model innovation and pragmatic engineering. He also designs data pipelines and semi-supervised labeling workflows, inventing an inexpensive Llama 2–based approach to validate noisy labels. Colleagues describe him as goal-oriented and detail-focused, equally comfortable shaping org-wide ML development standards and iterating on the low-level systems that make production at scale reliable.
14 years of coding experience
14 years of employment as a software developer
B.Eng./M.Sc. Computer Science, B.Eng./M.Sc. Computer Science at Universidad Pontificia de SalamancaAnyscale + Pinecone (https://www.anyscale.com/events/2024/03/08/rag-developer-bootcamp) RAG and LLMs, Anyscale + Pinecone (https://www.anyscale.com/events/2024/03/08/rag-developer-bootcamp) RAG and LLMs at RAG Developer BootcampCourse (268 hours) Business Administration and Economics, Course (268 hours) Business Administration and Economics at Community of MadridFSDL (https://fullstackdeeplearning.com/spring2021/) Deep Learning, FSDL (https://fullstackdeeplearning.com/spring2021/) Deep Learning at Full Stack Deep Learning 2021 EditionYahoo Leadership Program Organizational Leadership, Yahoo Leadership Program Organizational Leadership at The ForemCertificate of Training Advanced Scala, Certificate of Training Advanced Scala at Lightbend (formerly Typesafe)FSDL (https://fullstackdeeplearning.com/course/2022/) Deep Learning, FSDL (https://fullstackdeeplearning.com/course/2022/) Deep Learning at Full Stack Deep Learning 2022 EditionPh.D. Computer Science, Ph.D. Computer Science at Universidad Politécnica de MadridFDSL (https://fullstackdeeplearning.com/llm-bootcamp/) LLMs, FDSL (https://fullstackdeeplearning.com/llm-bootcamp/) LLMs at FSDL LLM Bootcamp
Spanish, English, French
Github Skills (101)
transformers10
hbase10
language-model10
python10
flax10
datasets10
deep-learning10
speech10
markdown10
transactional10
theory10
paper10
computer-vision10
transformer10
nlp10
Programming languages (9)
Github contributions (5)
francisco-perez-sorrosal/voodoo
Jul 2011 - Oct 2011
Contributions:43 commits in 2 months
francisco-perez-sorrosal/francisco-perez-sorrosal.github.io
Jun 2016 - Dec 2024
Contributions:98 pushes, 2 branches in 8 years 7 months
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.