Virgilio Kmetzsch Top-Executive
I’m a Senior Research Scientist working in AI/ML, focused on applying machine learning to solve complex challenges in biomedical research. With over a decade of experience in the tech industry, I've worked in software engineering, R&D management, and hands-on research in machine learning and computational biology.
Currently, I develop AI-driven models to accelerate drug discovery. I enjoy collaborating with teams of scientists, engineers, and clinicians to transform complex biological and chemical data into insights that guide therapeutic development.
.png)
Industry:
Medicine and Biology
Headquarters:
Paris, Île-de-France
Title:
Senior AI Research Scientist
Company:
Aqemia
Company Industry:
Biotechnology
Company Website:
aqemia.com
Company Linkedin:
linkedin.com/company/aqemia
Company Size:
11-50
Current Job Duration:
3 mos
Country:
France
Languages:
English, French, Portuguese
Work Experience:
Title: Senior AI Research Scientist
Company: Aqemia
Company URL: https://www.linkedin.com/company/aqemia/
Industry: Biotechnology
Company Size: 11-50
Location: Paris, Île-de-France, France
Country: None
Employment Type: None
Start At: 10-2025
Ends At: Present
Description: None
Title: Owkin
Company: Owkin
Company URL: https://www.linkedin.com/company/owkin/
Industry: Internet
Company Size: 201-500
Location: None
Country: None
Employment Type: None
Start At: 06-2023
Ends At: 09-2025
Description: I led impactful AI/ML-driven projects aimed at advancing biomedical research, while also mentoring a small team of scientists to support their career growth and development.
➢ Led a project investigating optimal data representations for spatial transcriptomics (ranging from PCA to graph neural networks), specifically targeting spatial domain identification tasks. Implemented representation models using Python libraries Scanpy, Squidpy, and PyTorch Geometric.
➢ Contributed to Owkin's AI-driven target discovery engine by leveraging multimodal patient data, biomedical literature, and clinical trials data to identify top-ranked, promising therapeutic targets per cancer indication.
➢ Led a scientific effort in disease progression modelling, using multimodal data to estimate future disease trajectories for unseen patients, supporting prognostic enrichment in clinical trials., I worked on a project focused on benchmarking different deep learning-based representations for bulk RNA-seq data, aiming to improve overall survival prediction and gene essentiality prediction.
➢ Researched multi-head architectures (one head per learning task) and developed a multi-head autoencoder in PyTorch, incorporating both unsupervised and supervised loss components.
➢ Co-designed and implemented a robust experiment pipeline using a nested cross-validation framework, ensuring consistent validation of model performance and reproducibility.
➢ Co-authored a journal paper detailing the methodology and findings of the project, contributing to advancing the field.
Title: PhD fellow in Machine Learning
Company: Inria
Company URL: https://www.linkedin.com/company/inria/
Industry: Research
Company Size: 1001-5000
Location: Paris Area, France
Country: None
Employment Type: None
Start At: 10-2019
Ends At: 09-2022
Description: None
Title: PhD fellow in Machine Learning
Company: Institut Du Cerveau – Paris Brain Institute
Company URL: https://www.linkedin.com/company/institutducerveau/
Industry: Biotechnology
Company Size: 501-1000
Location: Paris Area, France
Country: None
Employment Type: None
Start At: 10-2019
Ends At: 09-2022
Description: None
Title: Research internship in Deep Learning
Company: Inria
Company URL: https://www.linkedin.com/company/inria/
Industry: Research
Company Size: 1001-5000
Location: Grenoble Area, France
Country: None
Employment Type: None
Start At: 02-2019
Ends At: 07-2019
Description: None
Title: Datacom
Company: Datacom
Company URL: https://www.linkedin.com/company/datacom-teracom/
Industry: Telecommunications
Company Size: 201-500
Location: None
Country: None
Employment Type: None
Start At: 01-2014
Ends At: 04-2017
Description: Datacom has the largest R&D team in Latin America among telecom equipment manufacturers.
I was the project manager of DmOS, Datacom's operating system running on the company's switching and routing equipments.
➢ Oversaw multiple releases of a large software project (80+ engineers in 10+ geographically distributed teams).
➢ Established project performance metrics, communicated to stakeholders and coordinated actions to meet requirements.
➢ Organized issue tracking, including identification, prioritization, action assignment and root cause analysis, in collaboration with application support engineers and product managers., I was the manager of three geographically distributed teams (20+ engineers total), working on two product lines (SDH multiplexers and metro ethernet switches).
➢ Led the entire software development life cycle, from specification to deployment and maintenance.
➢ Supported engineers in developing their careers.
➢ Conducted the recruitment of 10+ team members., I lead the embedded software development team (10+ engineers) of the first SDH multiplexer (high speed data transmission equipment) developed in Brazil.
➢ Designed and developed (in C language) applications running on embedded Linux, for configuring, monitoring and testing SDH multiplexers.
➢ Implemented Linux device drivers for low-level tasks such as FPGA programming and NAND flash memory access.
➢ Provided technical training for new engineers in the team.
Education:
School: Sorbonne Université
Degree: Doctor of Philosophy - PhD, Computer Science
Activities: None
School: National School of Computer Science and Applied Mathematics of Grenoble
Degree: Master of Science in Informatics, Data Science
Activities: None
School: Electrical Engineering - UFRGS
Degree: Engineer's degree, Electrical Engineering
Activities: None
Headline:
PhD, Senior AI Research Scientist @ Aqemia | AI for Science