Paris, Île-de-France, France
Mistraling since 2026
👉 phacetlabs.com - the platform to easily deploy AI agents you control Phacet helps operations, finance, and product teams automate high-volume document workflows by deploying AI agents (Phacets) on their real business data — powered by LLMs, business rules, and Python, with built-in human oversight. Our process is fast, structured, and built for real-world impact: 1) Identify the right use case — We work with your team to find the highest-value AI opportunity (invoices, reconciliation, document matching…). 2) Deploy your first Phacet in 5 days — An AI agent live on your actual PDFs, spreadsheets or system exports. 3) Expand & scale — Launch more agents independently or with our support, while continuously improving performance. Phacet is trusted by teams who want the speed of LLMs without losing control, transparency, or context. 📩 Reach out to see how quickly you can automate your first process.
Ubble.ai is a company whose ambition is to revolutionize how we connect with digital services. We are solving the issue of securely validating consumers identity, by combining Video Streams, Computer Vision & Deep Learning, in a state-of-the art web-app, optimized for UX. Today the identity verification process (KYC) creates friction for the user and is generally not secured enough. Our ambition is to transform this moment into a super smooth and secured experience that will accelerate the onboarding process. Ubble.ai intends to offer a European alternative to GAFAs, to allow consumers to easily and securely confirm their online identity in a privacy-respectful way.
As a data scientist at Gamma, I had the opportunity to apply a wide range of algorithms in various industries. Main achievements include: - Applying both deep learning and non-learning computer vision models to automate certain tasks on french highways (incident detection, car model classification, plate recognition) - Building forecast and stock optimization models for two major fashion retailers - Contribute building a unique matching system in the diamond industry
Kayrros was founded in January 2016, by some of the best players of the Oil & Gas industry, backed by world-class mathematicians and data scientists. Ever since, Kayrros’ main purpose has been to make energy data accessible, usable, digestible and foremost actionable, to enable faster and better informed decisions. As one of the very first members of Kayrros’ team, I had the opportunity to work as product manager through every step of a product’s creation, from R&D to production. Key achievements includes: - Developing from scratch the product’s core algorithm using state of the art Natural Language Processing (NLP) technics, and reaching highly satisfying prediction results on specific financial trends - Bringing Kayrros’ first product to market