Pasteur Labs is a venture-backed public-benefit corporation that builds simulation intelligence technologies from theoretical foundations to reliable systems. It aims to facilitate research and development in various scientific domains via innovative simulation methodologies.

Pasteur Labs

Pasteur Labs &amp; ISI | 1,331 followers on LinkedIn. &quot;Disruptioneering&quot; for humanity&#39;s greatest challenges | Pasteur Labs &amp; Institute for Simulation Intelligence (ISI), reshaping the scientific method for the machine age.

Nikolas Borrel-Jensen

Principal Software Engineer

Youcef Rahal

AI & Simulation R&D 🐙

Diego Andrade

Senior Product Engineer at Pasteur Labs

Juan Pablo Carzolio

Pasteur Labs & ISI

When asked "Do your clients primarily use #RealWorldData , #SyntheticData , or a combination of both in their work? How do your tools handle each type, and how do you integrate them into your #MachineLearning #algorithms ?", our Founder Alexander Lavin shed some light on the engineering reality ( https://lnkd.in/e49B4Jav ): "That may be the #1 fault of existing technologies that we’re addressing: the inability to utilize real-world data. One cannot simply plug-in data—this is a fundamental flaw with CAE software, and it’s muddied with incumbents trying to throw ‘AI this’ and ‘chatbot that’ into the legacy CAE software toolchains... Industries are always naive about how hard it is to operationalize data in ways that build value — firehoses are expensive, and hard to tame in environments that matter... You’ll see in the upcoming Pasteur Labs platform launch, that we’ve built automated dataflow pipelines for interoperable #CAE toolsets and ML-based problem solving — that is, the necessary parameterization and labeling for algorithms (learning, optimization, and data assimilation), data structures befitting differentiable physics and interfaces for efficient, end-to-end modeling workflows." 👉 follow us Pasteur Labs & ISI (@SimAI4Science on https://lnkd.in/gznPy7Bh ) for more demos, product rollouts, and real-world results.

Important food-for-thought on #AI #ML advances vs #DigitalEngineering practice in this lecture! Here are a few key quotes from our R&D Lead Dion Häfner : 👨🍳 "Ideally a useful AI model, like we're building as plug-n-play commodities for industrial R&D, would have a few key ingredients for [human-machine teaming]: an understanding of *causality* s.t. can transfer understanding to us humans, *compatibility* for fusing your various models with mixed data end existing simulators is really attractive if not fundamentally necessary, and *reliability* cannot be an afterthought... or AI that speaks our language." 🔍 "thousands of nodes [in Deepmind's GraphCast] is hardly a small textbook example, where in the millions of nodes it gets interesting, so this SOTA in literature is several orders of magnitude away from commercial scale." 🕸 "Nvidia is all about Fourier Neural Operators and their kinda-platform called Modulus, which has inspiring science works, like Nvidia weather forecasting models based on FNOs, that is published and promoted but oops! They cannot work on unstructured meshes!" 🚀 "Pasteur ( pasteurlabs.ai ) is building a pure platform of AI surrogate pipelines for real-world engineering: it's critically important to validate surrogate models on actual data, that work in actual environments that are not cherry-picked, that interface with existing simulators... don't stop at training the models–that's the easy part! Integrating these models such that end-users can interact with them, that requires much more effort and thought, and is really unchartered territory..." If this fascinates and motivates you, check out #AI4Science #SimulationIntelligence opportunities with the Pasteur Labs team: https://lnkd.in/gz38vSg7 💪

It was an honor to join David Barnes, PhD and Marne Marotta today in a discussion on the militarization of AI and its implications for the future battlefield. I can't think of a more relevant topic given the acceleration of the so-called "AI arms race" and the rise of dual-use AI — all with Veteran's Day coming up and our recently-concluded presidential race in the backdrop. It was an incredibly rich conversation! We touched on topics ranging from how AI might reduce human error on the battlefield ... to cybersecurity (both using and targeting AI) ... to over-reliance on AI systems ... to AI training and awareness ... to how AI might impact the nature of war altogether ... and everything in between. Several themes emerged — my personal favorite being the importance of human-AI interaction design. Thank you to Arnold & Porter ’s Veterans and Affiliates Leadership Organization (VALOR) and to everyone who attended!

M. Alejandra Parra-Orlandoni (mapo)

The word is out! Turns out AI is NOT a magical solution to all the things. 😉 But what about AI for simulating physics in the context of industrial design and engineering? Seems like using AI for something science-y should be fine, right? Well, like most things, it's a bit more complicated than that. Check out this panel, hosted by nTop , to find out what it takes to use AI in in these contexts so that it is reliable, trustworthy, and effective. When: 31 Oct, noon EDT 👻 Register: https://lnkd.in/eQRz4_6H Bonus: Pasteur Labs & ISI CEO, Alexander Lavin , will be speaking on the panel!

AI + Physics + Simulation It's what we're all about at Pasteur Labs! Check out Alexander's talk at CDFAM, where he explains how we improve engineering workflows, and how this accelerates R&D in critical industries like energy, national security, transportation, and more! Thanks CDFAM - Computational Design Symposium for the fantastic event!!

Alexander Lavin of Pasteur Labs & ISI presented on "Realizing Differentiable Physics in Digital Engineering," at CDFAM - Computational Design Symposium in NYC, 2024. He introduced Pasteur Labs’ focus on simulation intelligence (SI)—integrating AI and advanced simulations to bridge the gap between what happens in silico and in reality. Their Auto Physics platform enables seamless workflows between CAD, CFD, and FEA tools with automation, cloud deployment, and machine learning-ready data. Lavin stressed that scaling and validating digital models with real-world feedback is essential to improve engineering outcomes and unlock new possibilities for advanced manufacturing and digital engineering. See the full presentation on Youtube: https://lnkd.in/etCuhJQf Subscribe for the latest as they are posted live from the archive.

CDFAM - Computational Design Symposium

Tomorrow's the day! Join me and three terrific women leaders — Usha Jagannathan, PhD , Renée Cummings , and Noelle R. — as we take the stage at #GHC24 ! The Grace Hopper Celebration is the world's largest gathering of women and nonbinary technologists from around the world. This year's theme is ME + WE, which perfectly captures how we must approach building technology if we want the outcomes to be empowering and inclusive. And on that note: Our panel discussion will cover all things #responsibleAI ! Our discussion will range from what worries us to what excites us about AI — and how we're navigating these dynamics as visionaries, pragmatists, and catalysts. We'll also share guidance on how to chart your own Responsible AI career path. 🤩 A special thanks and congrats to Usha for convening the panel and submitting the proposal! It's no small feat to be selected as one of ~30 feature events from thousands of submissions. Please join us in Philly! Register here https://ghc.anitab.org/ , and additional details below. Hope to see you there!! Where/when: Wednesday, Oct 9th | 1:00 PM - 1:45 PM EDT | PCC - 204ABC | Pennsylvania Convention Center, Philadelphia

🎧 https://lnkd.in/eusDfr2t 🎧 Hear from our Founder Alexander Lavin on, [🔊 9:20] why and *why now* for Simulation Intelligence [1], the unique operating principles of Pasteur Labs & ISI, and "the responsibility to build Nobel-Turing technologies that advance science & society for all humankind." [🔊 31:10] R&D software flywheels as a product, and building a "pure OS-type platform" [2]. [🔊 35:00] building product that "delivers AI-native physics value-add in a mostly automated way, a frictionless & interoperable way for our users, and a reliable & scientifically rigorous way... so our end-users can trust it is going to translate to the real world and faithfully represent the physics of the environments they're working in." [🔊 39:30] building the team that builds the IDE for Reality: "rather than 'retaining talent', my view is hiring well and then prioritizing their careers—which are best aligned with who we are and Pasteur Labs and what we're doing at Pasteur Labs, that's why they were hired... I prioritize the individuals and work to level them up, even if in the moment it deviates from the Pasteur roadmap needs, in the long-run that is the best talent retention policy." [🔊 50:00] *why* rather than how to start a company: "The entrepreneur aspect should be a consequence rather than a goal. Aspire to build. If what you're going to build requires entrepreneurship, do that. Aspire to invent technologies, not to have citations on google scholar. Aspire for the real-world instance of tangible impact, not the retweet metrics in these shallow reality bubbles... you need to value the less glamorous, not so shiny moments." [1] https://lnkd.in/eXKQxZBE [2] https://lnkd.in/e-vXSTHP

Our CEO, Alexander Lavin , took the stage at CDFAM - Computational Design Symposium today. He broke down how Pasteur Labs & ISI products can — not just solve many existing pain points — but also accelerate computational engineering and R&D. Fundamentally, we’re guided by product basics, for example: 👉 User-centric. Understanding the problems and pain points of our users, like low efficiency or high cost of running simulations. 👉 Right (AI) tool for the job. Using SciML (scientific machine learning) to provide continuous and differentiable data to engineers, which unlocks tons of design & analysis capabilities often not available today. 👉 Interoperable. Making it easy to integrate with existing tech stacks and a variety of tools involved in computational engineering and R&D. I’m so excited — and driven — by the opportunity for our tools to unlock innovations in energy, transportation, medical devices, national security, and more! Peter Sharpe Alex Kaszynski Nikolaos Bouklas

Awesome intro to what we do @ Pasteur Labs & ISI ! In short: "What if that sim-to-real delta could be optimized to zero by making digital engineering with differentiable software?" https://lnkd.in/dbDAUqYa

PAST

AI:PASTEURLABS

Pasteur Labs