Sponsored LinkedIn post
A dedicated post to introduce your brand to my audience.
Master Thesis Researcher, Discrete Diffusion Models for Structured Sequence Reconstruction @ CentraleSupélec
Companies pay Yacine to write a post about their product on LinkedIn.
This page is where they book it. Below: Yacine's audience, recent posts and price. Naano, the platform behind this page, handles the brief, your approval of the post and the payment.
€113per post, fixed price
or 3 posts for €319 · You approve the post before it goes live
Looking for Yacine? View LinkedIn profileA creator yourself? Create your own page
MSc Researcher @ CentraleSupélec | Incoming Founder @ Entrepreneurs First
MSc Engineering student at Polytech Nancy – Université de Lorraine, currently completing my Master's thesis at CentraleSupélec on discrete diffusion models for DNA sequence reconstruction from noisy sequencing reads (publication in preparation). My research sits at the intersection of generative models, structured sequence data, and scientific applications. I build things from scratch — from Transformer architectures and diffusion training pipelines to full genomic data generation tools — because understanding every layer of the stack is how I do research. Current work: • TraceDiff — Discrete diffusion architecture for reconstructing corrupted DNA sequences via iterative denoising (CentraleSupélec / IETR, UMR CNRS 6164) • SynthDNA-Lab — Open-source synthetic DNA dataset generator with 7 vendor profiles and 3 sequencing backends (Illumina, Nanopore, PacBio): github.com/yacine-baghli/SynthDNA-Lab • MiniLM-Bench — 7 Transformer attention mechanisms implemented and benchmarked from scratch on 1B tokens: github.com/yacine-baghli/minilm-bench Recent projects: • KAIROS — AI-assisted genomic case triage (SOPHiA Genetics × ETH Zurich Hackathon) • World Models — JEPA-based representation learning (preparing "Hack the Worlds", Paris) • Multimodal Embodied Agent with cross-session memory (HACK 2026, ETH Zurich) • 2nd Place — EPFL Quantum Hackathon, comparison between QML and ML approaches for option pricing Self-directed coursework: MIT 6.S191 (Deep Learning), Stanford CS336 (Language Modeling from Scratch), Stanford CS229 (ML / LLMs), Stanford CS230 (Deep Learning), Stanford CME296 (Diffusion & Large Vision Models), Stanford CME295 (Transformers & LLMs). Python · PyTorch · CUDA · C/C++ · Mixed-Precision GPU Training
My most engaged audience
53% Founder / C-level · 68% Engineering / Data
From 235 people who reacted to or commented on my posts
Public LinkedIn data · Sep 17, 2026
🏆 We just won 1st place at the OpenAI Build Week Hackathon in Paris This weekend, we spent the day building with Codex and walked away with first place. We designed Mimex a tool for Codex that mimics your actions to learn a task, just by watching you do it once. → The goal: let anyone hand a task to an agent without writing a single line of instructions, just show it. How it works: You record yourself completing a task on any website, talking through it out loud as you go. Mimex turns that recording into a skill md file, then uses that skill to execute the task for you. Show it once. It learns. It does. What makes it different: It adapts to any interface, no brittle scripts, no hardcoded selectors It works on any site It handles the unexpected, because it understands the context of what you showed it, it can deal with variables and situations it was never explicitly walked through Voice instructions layer intent on top of action, so the agent knows the why, not just the what We even shipped it to Ginse (an app store for agents), so it's live and ready to use. But what made this experience truly exceptional wasn't just the system. It was the people. A room full of builders at 42 Paris's campus, all pushing raw ideas into working demos in a single day. These are the environments where you don't just build fast, you learn how to think better. Huge thanks to the team who made this happen: Oscar HU, Adrien Rouland-Devienne and Benjamin Merchin. This one's ours. 🙌 And massive thanks to Boris de Wit and Pierina Camarena for putting together an outstanding event. Special thanks to Gibril Gabriel Bachouchi and Slava Mikhalkin for their invaluable advice, and to the whole judging panel for the sharp feedback and the vote of confidence. Grateful for the intensity, the conversations, and everything I'm taking away from this weekend. #OpenAI #Codex #AIAgents #Hackathon
Happy to share that I'm joining The Bridge, Entrepreneurs First's fall cohort in San Francisco. EF backs exceptional people at the very beginning. The companies built that way are worth north of $18B today. Forty of us from Europe, one house in San Francisco, eight weeks to start. The teams and the companies get built there. I'm arriving with two claims: > Biology is the next computing substrate. We finally understand enough of the mechanism for that to stop being a metaphor. > The right architecture beats another order of magnitude of data. Scaling assumes data is cheap and abundant. Observation scales. Intervention does not. If you're a founder in SF working on bio, agentic systems, or robotics, I want to talk. And if you're in none of those but you hold a bold claim you're willing to defend against someone who'll push back on it, I especially want to talk. Those conversations are the whole reason I'm getting on the plane. Comment or DM. Coffee's on me. Thanks Matteo, Anastasia and the team for the conversations that got me here.
Europe has some of the best builders in the world. You can meet them this weekend. Over the past year I've shipped 7 products at hackathons across Europe: first place at the OpenAI Build Week hackathon in Paris, second at QPFL, the Google DeepMind track at RAISE, Hack2026 at ETH Zurich. Every one of those weekends, I walked out with people I'm still building with today. But almost nobody finds these rooms. Different cities, different languages, same pattern every single time: brilliant builders doing extraordinary work, invisible to the equally brilliant builders three countries away. Which is why I've joined Hackathon Hub as an Ambassador. At Hackathon Hub we're building one place for Europe's hackathon ecosystem: events that are easier to discover, communities connected across borders, scattered local scenes turned into one network. The platform launched barely a month ago and there are already 110+ ambassadors and more than 2,000 builders following along, which tells you how badly this was missing. My focus is turning hackathons into real launchpads. The best hackathons don't end with a prize, they end with a door opening: an intro, an interview, a programme that takes you seriously. That's the kind of event I want to help build in Europe. If you go to hackathons, every event in Europe is now in one place. No more finding out about the good ones after they're over. And if you organise them or run a community, bring it into the network. We're looking for ambassadors in every city. Either way, reach out. Link in the comments. 🇪🇺 #Hackathons #AIAgents #Europe #Building
Choose a format for your next LinkedIn collaboration.
A dedicated post to introduce your brand to my audience.
A series of posts to build visibility over time.
A series of posts to build visibility over time.
Your collaboration, managed on Naano
How does this price compare? B2B creator rates 2026
Based on 235 people who reacted to or commented on my posts. Percentages use the profiles that could be classified.
Based on 235 people who reacted to or commented on my posts. Percentages use the profiles that could be classified.
Based on 235 people who reacted to or commented on my posts. Percentages use the profiles that could be classified.
Based on 235 people who reacted to or commented on my posts. Percentages use the profiles that could be classified.
Storefront by Naano · Create yours
Yacine Baghli is a B2B LinkedIn creator based in France with 5,589 followers, posting about AI, HealthTech, and Developer Tools. On Naano, one sponsored LinkedIn post by Yacine Baghli costs €113, a fixed price per post.
The Naano team casts the creators, writes the briefs, reviews every draft with you and reports results per creator. You approve the shortlist and every draft.