Franz Cesista
Metro Manila, Philippines
Summary
🤩
Rockstar🎓
Top SchoolFranz Cesista is a machine learning research engineer with 10 years of experience building production-grade AI systems for document information extraction in logistics. At Expedock he architected end-to-end ML infrastructure—training, deploying, and monitoring hundreds of multimodal models on Triton/SageMaker and a Snowflake/DBT-backed data stack—enabling high-throughput, low-cost inference on a single A10G. He has deep systems-level expertise, from C++ implementations of model building blocks (including a LlamaV2 implementation) to reducing PyTorch/CUDA memory usage so more inference jobs run per GPU. A mathematician by training and a two-time IOI and two-time ICPC World Finalist, he combines rigorous algorithmic thinking with pragmatic engineering. He also has product-facing experience shipping data products and visualization tooling that turn parsed documents into actionable insights for logistics executives. Notably, he discovered unusual structure in real datasets (a torus-like manifold) and enjoys optimizing ML stacks at every level of abstraction.
10 years of coding experience
6 years of employment as a software developer
High School Diploma, High School Diploma at Philippine Science High School - Eastern Visayas CampusBachelor of Science - BS, Mathematics, Bachelor of Science - BS, Mathematics at Ateneo de Manila University
Filipino, English, German
Github Skills (111)
transformers10
language-model10
python10
flax10
prompt-engineering10
deep-learning10
validation10
symbolic-ai10
generative-ai10
ctc10
transformer10
transfer-learning10
nlp10
jax10
json10
Programming languages (8)
Github contributions (5)
gdsc-loyola/amsearch
Sep 2019 - Sep 2022
A search site for the past Long Tests and Finals collected by the Ateneo Math Society's Academic Enrichment Team
Contributions:3 PRs, 8 pushes, 1 branch in 3 years
finalsmathsite-searchenrichmentlong
leloykun/NOI-Practice
Nov 2016 - Feb 2018
Contributions:125 pushes, 1 branch in 1 year 3 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.