Yannis Flet-berliac
Research Scientist at Cohere
Paris, Ile-de-France
Summary
👤
Senior🎓
Top SchoolYannis Flet-berliac is a research scientist with 10 years of experience advancing reinforcement learning and language model research, currently focusing on post-training and RL for LLMs at Cohere. His trajectory ranges from a PhD at Inria on sample-efficient deep RL for control, exploration and safety to a Stanford postdoc in Emma Brunskill’s lab where he translated offline and imitation learning methods to robotics, healthcare and education. He has blended industry and research roles—building NLP systems as an ML engineer, applying RL and LLMs at InstaDeep for genomics, and contributing to computer vision and speech projects across startups. Yannis combines rigorous theoretical grounding with practical deployment experience, routinely moving algorithms from papers to production-ready systems. He also brings academic service leadership, having served as Associate Program Chair for ICML, reflecting a strong connection to the research community. Based in Paris, he speaks the language of both cutting-edge research and real-world engineering, with an uncommon breadth across modalities (NLP, CV, speech) and application domains.
10 years of coding experience
4 years of employment as a software developer
Postdoc, Computer Science, Postdoc, Computer Science at Stanford UniversityClasses Préparatoires aux Grandes Écoles (PCSI and PC*), Classes Préparatoires aux Grandes Écoles (PCSI and PC*) at Lycée Pierre De FermatDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Université de LilleEngineering, General Engineering Diploma (MSc), Engineering, General Engineering Diploma (MSc) at Centrale NantesTechnical University of DenmarkMaster's degree, Management of Digital Business & IT, Master's degree, Management of Digital Business & IT at Audencia
French, English, Spanish, Danish
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (66)
bots10
botkit10
dialogue-systems10
assistants10
voice10
conversational-ai10
dialogflow10
nlp10
chatbot10
conversation10
google-assistant10
reinforcement-learning-environments10
reinforcement-learning10
bot-framework10
watson10
Programming languages (4)
Github contributions (5)
yfletberliac/actor-with-variance-estimated-critic
Mar 2020 - Oct 2020
Is Standard Deviation the New Standard? Revisiting the Critic in Deep Policy Gradients (TensorFlow).
Contributions:1 review, 7 commits, 1 PR in 6 months
policygradientsdeviationpolicy-gradientsdeep-learning
yfletberliac/rlss-2019
May 2019 - Aug 2019
[RLSS 2019] Bandits, RL & Deep RL: Practical Sessions.
Contributions:124 commits, 2 PRs, 62 pushes in 2 months
practical-sessionsreinforcement-learningdeep-reinforcement-learningpracticaldeep-rl
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.