Umberto Griffo
Senior MLOps Engineer at Promaton
Lisbon, Lisbon, Portugal
Summary
👤
Senior🎓
Top SchoolUmberto Griffo is a Staff MLOps/ML Engineer with 13 years of experience building and operationalising machine learning and data platforms across enterprise and regulated medical environments. He has a strong track record of productionising ML/DL models—most notably reducing GPU inference latency and memory by ~50% for 3D computer vision pipelines used in automated surgical planning. Equally comfortable with large-scale batch and real-time data engineering, he’s implemented robust pipelines using Spark, Kafka, Flink and cloud-native AWS tooling, and designed MLOps frameworks to support data science teams at scale. His hands-on work spans deep learning for computer vision, predictive maintenance, and vector-store backed AI agents (refactoring backend memory and FAISS/Qdrant integrations). Umberto combines technical leadership—authoring ADRs/RFCs and setting engineering quality standards—with mentoring and cross-functional alignment between ML, frontend and product teams. Based in Lisbon with a master’s in Computer Engineering, he’s currently focused on operationalising models on Kubernetes with Ray Serve, bringing pragmatic performance-first improvements to production ML.
13 years of coding experience
10 years of employment as a software developer
Bachelor's degree, Computer Engineering, Bachelor's degree, Computer Engineering at Università degli Studi di Roma 'La Sapienza'Hadoop, Hadoop at Cloudera University
Stackoverflow
931reputation
85kreached
19answers
0questions
Github Skills (15)
vector-search10
agent10
faiss10
python10
ai10
langchain10
back-end-development10
dockers7
docker7
hive6
apache-spark6
feature-selection6
dataframe6
scala6
java6
Programming languages (3)
Github contributions (5)
cheshire-cat-ai/core
Mar 2023 - Mar 2023
AI agent microservice
Role in this project:
Back-end Developer
Back-end DeveloperContributions:9 commits, 6 PRs, 10 comments in 7 days
Contributions summary:Umberto primarily focused on refactoring and enhancing the back-end functionality of the AI agent microservice. They implemented improvements and configurations for the vector store, initially using Qdrant and later transitioning to FAISS. Additionally, they modified the code, including the `memory.py`, `looking_glass.py` and `rabbit_hole.py` files, and adjusted file handling to better integrate the vector store and enhance the overall system design. These changes improved the agent's memory and its ability to ingest and retrieve information.
agentaiassistantbotbot-framework
umbertogriffo/umbertogriffo
Dec 2020 - Jan 2025
Contributions:38 pushes, 1 branch in 4 years 2 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.