May 2024 – Present · 2 yrs · Paris, Île-de-France, France · On-site

• Bringing AI agents to life one commit at a time
• Design system
• Infrastructure
• Agent observability
• Organizer of the "Engineering Night" events gathering up to 200 engineers around technical talks

Aug 2023 – May 2024 · 9 mos · Paris, Île-de-France, France · On-site

• Lead AI R&D (RecSys, MaaS, NLP)
• Built the AI infrastructure from the ground up
• Stack: Python (HuggingFace ecosystem, FastAPI, Lightning), K8, Helm, Temporal, Redis, Postgres, MaaS

Jan 2023 – Aug 2023 · 7 mos · Paris, Île-de-France, France · Remote

• We created 𝐋𝐚𝐧𝐠𝐓𝐞𝐬𝐭; a library for people to deliver safe and effective language models.
• Spearhead the development of 𝐋𝐚𝐧𝐠𝐓𝐞𝐬𝐭, a powerful library designed to elevate the reliability and effectiveness of language models.
• Lead a cross-functional team of 5 engineers
• Craft pioneering MLOps workflows using LangTest for benchmarking leading NLP frameworks.
• Drive the integration of top-tier NLP libraries including transformers, sparknlp, langchain, and spacy.

Oct 2020 – Feb 2023 · 2 yrs 4 mos · Copenhagen, Capital Region of Denmark, Denmark · Hybrid

• Engineered multilingual NLP solutions tailored to process real-world data for applications such as sentiment analysis and semantic search.
• Executed end-to-end MLOps strategy, handling everything from data annotations to model deployment and monitoring, achieving 1B transformer-based models inferences/day on a Kubernetes cluster.
• Oversaw a dedicated team consisting of computational linguists and data annotators
• Stack: Python (PyTorch / Transformers / SpaCy), ONNX, GCP, Scala, Docker, K8s, ElasticSearch

Jun 2019 – Jul 2020 · 1 yr 1 mo

• Deep learning and Machine Learning algorithms for NLP. Data extraction / Modelling / Model assessment / Model deployment and scaling / MLOps
• Key achievement: fine tuning BERT for a multi-label classification task + deployment at scale (~300req/sec).
• Stack: Python (Sklearn, PyTorch, Transformers, SpaCy, Flask, BeautifulSoup) / Google Cloud Platform / IBM Watson / Elasticsearch / Docker / CircleCI

Mar 2019 – Jun 2019 · 3 mos

• Machine Learning algorithms for NLP.
• Stack: Python (PyTorch / Sklearn / SpaCy / BeautifulSoup) / Google Cloud Platform

Jan 2020 – Jul 2020 · 6 mos · Copenhagen Metropolitan Area

• Research Topic: 𝐀𝐧𝐨𝐦𝐚𝐥𝐲 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐈𝐧 𝐇𝐢𝐠𝐡 𝐃𝐢𝐦𝐞𝐧𝐬𝐢𝐨𝐧𝐚𝐥 𝐓𝐢𝐦𝐞 𝐒𝐞𝐫𝐢𝐞𝐬 𝐮𝐬𝐢𝐧𝐠 𝐃𝐞𝐞𝐩 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠
• Implemented SOTA papers using the following techniques: Autoencoders / Attention mechanisms / LSTMs / Forecasting / Clustering algorithms / Anomaly severity detection
• Manipulated large amount of data (~5TB) using Kusto
• Stack: Python (PyTorch / Sklearn) / SQL

Apr 2018 – Aug 2018 · 4 mos · station f

• Back-end development of a platform counting more than 20000 users. Setting up of a matching algorithm based on users' preferences and information. Optimisation of database structure and query system.
• Stack : Symfony / SQL