October 31 - November 1 - Co-Located Events
October 28-30 - Conference
Lyon Convention Centre - Lyon, France
More information for Open Source Summit + Embedded Linux Conference Europe 2019

Sign up or log in to bookmark your favorites and sync them to your phone or calendar.

LF AI Summit (AI/ML/DL) [clear filter]
Monday, October 28


Machine Learning Models and Datasets Versioning Practices and Tools - Dmitry Petrov & Ruslan Kuprieiev, Iterative AI
The rise of AI and ML changes development workflow and requires new development tools: data versioning, ML pipeline versioning, experiments metrics tracking and others that have not been formalized and even named yet.

Machine learning workflow is data-centric in contrast to source code-centric software engineering workflow. The traditional software engineering toolset does not fully cover ML team's needs. We will discuss the current practices of organizing ML projects using traditional open-source tools like Git and Git-LFS as well as their limitations. Thereby motivation for developing new ML specific data management systems will be explained.

Data Version Control or DVC.ORG is an open source, command-line tool. We will show how to version datasets with dozens of gigabytes of data and version ML models, how to use your favorite cloud storage (S3, GCS, or bare metal SSH server) as a data file backend and how to embrace the best engineering practices in your ML projects.

avatar for Dmitry Petrov

Dmitry Petrov

Co-Founder & CEO, DVC
Dmitry is an ex-Data Scientist at Microsoft with Ph.D. in Computer Science and active open source contributor. He has written and open sourced the first version of DVC.org - machine learning workflow management tool. Also he implemented Wavelet-based image hashing algorithm (wHash... Read More →
avatar for Ruslan Kuprieiev

Ruslan Kuprieiev

Software Engineer, Iterative AI
Ruslan is a Software Engineer at Iterative AI. Previously he worked on live container migration at Parallels, Linux Kernel live-patching at CloudLinux, and also in a few startups. Ruslan's career started by working in an open source project called CRIU and he continues to contribute... Read More →

Monday October 28, 2019 11:30 - 12:05
St. Clair 3
  • Session Slides Included Yes


Lightning Talk: A Perfect Match: AI4EU and Acumos for Europe - Martin Welss, Fraunhofer IAIS
This presentation will make the link between Linux Foundations' Acumos project and the AI4EU project, which has started at the beginning of 2019 and will run for 3 years. At the beginning of 2020 we will present to the public the first version of the AI4EU platfrom that already includes an Acumos instance with onboarded AI Tools. The AI4EU platform will make later on easily accessible to the community a lot of very interesting datasets like satellite images or from the human brain project in conjuction with selected and proven AI models and tools on the AI4EU Acumos instance. The goal is to raise attention to the project and gather feedback from the community.

avatar for Martin Welss

Martin Welss

Research Engineer, Fraunhofer IAIS
Martin has more than 25 years expierence of developing and coordinating Java Enterprise Applications on Linux. He now works for Fraunhofer IAIS on the AI4EU project, that belongs to the EU Horizon 2020 program to improve AI acceptance across Europe. He specifically works on the integration... Read More →

Monday October 28, 2019 12:20 - 12:30
St. Clair 3
  • Session Slides Included Yes


Lightning Talk: Using Data without Compromising Privacy - Gordon Haff, Red Hat
Deep learning and machine learning more broadly depend on large quantities of data to develop accurate predictive models. In areas such as medical research, sharing data among institutions can lead to even greater value. However, data often includes personally identifiable information that we may not want to (or even be legally allowed to) share with others. Traditional anonymization techniques only help to some degree.

In this talk, Red Hat's Gordon Haff will share with you the active research activity taking place in academia and elsewhere into techniques such as multi-party computation and homomorphic encryption. The goal of this research is to enable broad information sharing leading to better models while preserving the anonymity of individual data points.

avatar for Gordon Haff

Gordon Haff

Writer, opensource.com
Gordon Haff is Technology Evangelist at Red Hat where he works on emerging technology product strategy; writes about tech, trends, and their business impact; and is a frequent speaker at customer and industry events. Among the topics he works on are edge, blockchain, AI, cloud-native... Read More →

Monday October 28, 2019 12:30 - 12:40
St. Clair 3
  • Session Slides Included Yes


The Twilight Zone: The AI / Privacy Governance Dilemma - Maru Rabinovitch, Maru Rabinovitch Legal
Artificial Intelligence (AI) is becoming an important component of most business models. As a result of this technology’s rise and the data collection and use AI requires, many companies find themselves facing new and complicated privacy and ethical governance issues. In this presentation, Maru Rabinovitch, will use specific examples to illustrate concerns that arise from the use of AI without proper governance; discuss privacy and ethical governance considerations; and address some of the enacted and pending regulations that may affect the use of AI.

avatar for Maru Rabinovitch

Maru Rabinovitch

Privacy & Open Source Attorney, Maru Rabinovitch Legal
Maru Rabinovitch is a Certified International Privacy Professional (CIPP/US), who graduated with a Privacy Certificate from law school. Her areas of practice include commercial transactions, privacy, technology/open source licensing, and trademark protection. During the past twelve... Read More →

Monday October 28, 2019 14:25 - 15:00
St. Clair 3
  • Session Slides Included Yes


Are Data Struggles Holding Back your AI Projects? Are you Ready for Open Metadata and the CDLA? - Jeffrey Borek & Mandy Chessel, IBM
Data challenges are halting AI projects for multiple reasons, and open source developers are looking for solutions. Do you know how to share data sets properly? Just like software, you don't want to put your data sets out in the public domain without proper license protections. The Community Data License Agreement (CDLA) is a key part of the answer.

About 80% of the work with an AI project is collecting and preparing data. Are you having challenges with 'data sprawl' across your company? How about GDPR compliance? An open metadata strategy can help. Open source project Egeria provides the open metadata and governance type system, frameworks, APIs, event payloads and interchange protocols to enable tools, engines and platforms to exchange metadata. Leading project community members bring experience from their roles at HortonWorks, IBM, Index Analytics, ING, SAS, and others. Come join this session to learn how to get the best value from data whilst ensuring it is properly governed.

avatar for Mandy Chessel

Mandy Chessel

ODPi TSC Chairperson and ODPi Egeria project chairperson. IBM Distinguished Engineer, IBM
Mandy Chessell CBE FREng CEng FBCS is an IBM Distinguished Engineer, Master Inventor and Fellow of the Royal Academy of Engineering. Mandy is a trusted advisor to executives from large organisations, working with them to develop their strategy and architecture relating to the governance... Read More →
avatar for Jeffrey Borek

Jeffrey Borek

WW Program Director, IBM
Jeffrey Borek is a senior technology and communications professional with over twenty years of leadership and technical experience in the Software, Telecommunications, and Information Technology industries. He is currently the leader of the OSPO at IBM, and works in the Open Technologies... Read More →

Monday October 28, 2019 15:15 - 15:50
St. Clair 3
  • Session Slides Included Yes


AI/ML Deployment at the Edge - Andrea Gallo, Linaro
Arm and Linaro launched the AI initiative one year ago to collaborate on an open source inference engine common to all Arm edge devices and support SoC specific NN acceleration via a plug-in back end framework.

The mlplatform.org platform hosts the upstream open source work for both Arm NN and the Arm Compute Libraries. The team, made up of engineers from Arm, Linaro, Qualcomm, TI, and other members, is deploying Arm NN and the Arm Compute Libraries in edge devices with support for the most widely used frameworks like Tensorflow, ONNX, etc.

Andrea Gallo, Linaro VP of Membership Development, will provide an overview of the ongoing activities to support multiple SoCs in Arm NN, set up CI and testing infrastructure, integrate in runtime frameworks and graph compilation technologies.

avatar for Andrea Gallo

Andrea Gallo

VP of Membership Development, Linaro
Andrea joined the Linaro Technical Steering Committee in 2010 as an ST-Ericsson Fellow before becoming a Linaro employee in 2012 as the Director of the Linaro Enterprise Group (now known as the Linaro Data Center and Cloud Group). He then went on to work as the VP of Segment Groups... Read More →

Monday October 28, 2019 16:20 - 16:55
St. Clair 3
  • Session Slides Included Yes


Model as a Service for Real-time Decisioning​ - Sumit Daryani & Ravi Dubey, Capital One
Imagine a stream processing platform that leverages ML models and requires real-time decisions. While most solutions provide tightly coupled ML models in the use case, these may not offer the most efficient way for a data scientist to update or roll back a model. With model as a service, disrupting the flow and relying on technical engineering teams to deploy, test, and promote their models is a thing of the past. It’s time to focus on building a decoupled service-based architecture while upholding engineering best practices and deliver gains for model operationalization.

Sumit demonstrates a reference architecture implementation for building the set of microservices and lay down, the critical aspects of building a well-managed ML model deployment flow pipeline that requires validation, versioning, auditing, and model risk governance. See the benefits of breaking the barriers of a monolithic ML use case by using a service-based approach consisting of features, models, and rules.

avatar for Ravi Dubey

Ravi Dubey

Director, Software Engineering, Capital One
Ravi is a Lead Software Engineer, Team Lead/Architect and Director at Capital One specializing in Decision Processing, Platform Delivery, and Cloud Engineering.
avatar for Sumit Daryani

Sumit Daryani

Manager/ Architect Software Engineering, Capital One
Sumit Daryani is a software engineering manager and architect at Capital One. He works on a real-time machine learning decision platform to protect its banking platform and foster quick decisions to support the fraud strategy. Previously, Sumit was a full-stack engineer on a diverse... Read More →

Monday October 28, 2019 17:10 - 17:45
St. Clair 3
  • Session Slides Included Yes
Tuesday, October 29


Using Kubeflow Pipelines for Building Machine Learning Pipelines - Yufeng Guo, Google
Kubeflow is an open-source project dedicated to making deployments of machine learning workflows on Kubernetes simple, portable and scalable. This session will focus on Kubeflow Pipelines, a platform to enable end-to-end orchestration of ML pipelines as well as easy experimentation and re-use. You'll learn how to build and manage machine learning workloads that can scale.

Kubeflow is a very exciting open-source project that bridges the gap between the DevOps world with the machine learning world. There are many concepts that can be highly valuable to cross-pollinate between these worlds, and Kubeflow helps codify that into best practices.

Learn more about Kubeflow Pipelines at https://www.kubeflow.org/docs/pipelines/pipelines-overview/

avatar for Yufeng Guo

Yufeng Guo

Developer Advocate, Machine Learning, Google
Yufeng is a Developer Advocate at Google focusing on Cloud AI, where he is working to make machine learning more understandable and usable for all. He is the creator of the YouTube series AI Adventures, at yt.be/AIAdventures, exploring the art, science, and tools of machine learning.He... Read More →

Tuesday October 29, 2019 11:30 - 12:05
Rhone 1
  • Session Slides Included Yes


Ethics in AI: Detecting and Remediating Bias in AI by Creating Ethical AI Practices - Romeo Kienzler, IBM
One of the most critical and controversial topics around artificial intelligence centers around bias. As more apps come to market that rely on artificial intelligence, software developers and data scientists can unwittingly inject their personal biases into these solutions. In addition, the datasets curated over a period of time based on historical data have become inherently biased towards a particular gender, race and other attributes.

Given how these AI systems are utilized to make decisions in criminal systems, approve or deny college admissions, loans etc, it has become critical to have tools to detect and remediate these biased AI systems. We have launched AI Fairness 360, an open source library to detect and remove bias in models and data sets, with 70+ Metrics and 10 Algorithms.

We will share lessons learnt while using AI Fairness 360 and show how to leverage it to detect and de bias models during pre-processing, in-processing, and post-processing.

avatar for Romeo Kienzler

Romeo Kienzler

Chief Data Scientist, IBM

Tuesday October 29, 2019 12:20 - 12:55
Rhone 1


How Linux Foundation is Changing the (Machine-learning) World! - Natarajan Subramanian, Tech Mahindra
Open-source AI tools/solutions ARE available but they’re not easy to implement, aren’t always compatible, and each solve only a small piece of the puzzle. That’s why – despite growing adoption – AI is still difficult to deploy. That’s also why LF Artificial Intelligence Foundation (LFAI) was established – to reduce solution fragmentation, encourage project, company & developer collaboration, and drive the effective use of AI tools/solutions to increase adoption/innovation. LFAI ground-breaking projects include Acumos AI (open-source marketplace for Machine-Learning models initiated by ATT & Tech Mahindra) and Horovod, (distributed training framework for TensorFlow, Keras, & PyTorch contributed by Uber). Here Natarajan Subramanian explores LFAI projects & activities, including a new (very cool) AI open-source landscape tool. He also presents the opportunities and benefits of actively participating in the LFAI community

avatar for Natarajan Subramanian

Natarajan Subramanian

Head of Enterprise Architecture - AI, Digital & Cloud, Tech Mahindra Americas Inc
As a Head of Enterprise Architecture – AI, Digital & Cloud at Tech Mahindra (Americas) Inc, Natarajan (Nat) Subramanian is focusing on AI open source development and strategy. He is leading the architecture, development and technology strategy team. He is a LF AI foundation Governing... Read More →

Tuesday October 29, 2019 14:25 - 15:00
Rhone 1
  • Session Slides Included Yes


OpenSource AIOps - How to Kickstart your Journey - Marcel Hild, Red Hat
As IT operations become more agile and complex, at the same time the need to enhance operational efficiency and intelligence grows. The term AIOps is often mentioned in this context, as it promises to take operations to the next level. But attaching an AI system to your data center might be daunting. Fear not - you don’t need a full-fledged commercial product to start your journey.

Marcel will talk about the story, how Red Hat has experimented with operational data and ran small data science experiments on it. The good ones have been containerized and connected to production systems in our kubernetes clusters. Hear about the lessons learned and the tooling we’ve used.

And you start right away and use the same containers and opensource software in your environment. Amongst others, you will see anomaly detection for logs and load prediction with Prometheus metrics.

avatar for Marcel Hild

Marcel Hild

Engineering Manager, AI CoE, CTO Office, Red Hat
Marcel Hild has 25+ years of experience in open source business and development. He co-founded a Linux consulting company, worked as a freelance developer, a Solution Architect for Red Hat and core Developer for Cloudforms, a Hybrid Cloud Management tool. Now he researches the topic... Read More →

Tuesday October 29, 2019 15:15 - 15:50
Rhone 1
  • Session Slides Included Yes


Just Deploy It! How to Ship Your ML Model to Production - Marianna Diachuk, Women Who Code Kyiv
Once upon a time in the kingdom of Artificial Intelligence, there were data scientists who worked hard on their complex researches and ML models development. But then the day of atonement came - time to deploy the models to production. And at first, it seemed that everything they’ve been building so hard started to fall apart slowly piece by piece. But as data scientist read more manuscripts they’ve discovered there are many ways to ship your model using the open-source library with not much effort to be applied.
This talk will uncover the ways to ship the model to production - automated as well as potential custom solutions and also tools and libraries which can ease this process.

avatar for Marianna Diachuk

Marianna Diachuk

Data Scientist, Women Who Code Kyiv
- Leading a small but proud team of 2 data scientists and 1 data engineer. - Developed and deployed 7 models and shipped 26 models prototypes. - Organizing Women Who Code data science-themed events as Data Science Lead for 2.5 years - Presented 4 tech talks to my colleagues and the... Read More →

Tuesday October 29, 2019 16:20 - 16:55
Rhone 1
  • Session Slides Included Yes


GNES: An Opensource Generic Neural Elastic Search Framework for Searching Everything - Dr. Han Xiao, Tencent AI Lab
Tencent receives tons of text, images and videos everyday. Searching efficiently and effectively means everything to us, and understanding the content is the key to improve the search accuracy. Recent advances in deep learning (VGG/ELMO/BERT) allow one to uniformly represent the content using a dense vector regardless its form (text/video), which forms the backbone of our GNES. But GNES is more than a collection of popular algorithms, it provides an end-to-end solution optimized for user experience, enabling one to easily index, query multiple data types including text and other multimedia formats. Architecture-wise, GNES is an "all-in-microservice" solution that can be easily scaled on cloud services. Apart from the technical highlights, GNES is one of few projects at Tencent that is opensource from the day one. It follows the best-practice outside-in and creates a collaborative culture inside-out. As the lead of GNES, I will share the design principle and lessons learned with you.


Dr. Han Xiao

Engineering Lead, Tencent
Dr. Han Xiao is an Engineering Lead at Tencent AI Lab, a board member at LF AI Foundation, and the Chairman of the German-Chinese Association of Artificial Intelligence. Han received his Ph.D. and M.Sc. in computer science from the Technical University of Munich in Germany. At Tencent... Read More →

Tuesday October 29, 2019 17:10 - 17:45
Rhone 1
  • Session Slides Included Yes
Wednesday, October 30


Beaming Deep Learning with Ludwig - Suneel Marthi, AWS
Ludwig is a code-free deep learning toolbox based on TensorFlow open-sourced by Uber AI Labs. Ludwig is unique in its ability to help make deep learning easier to understand for non-experts and enable faster model development cycles for experienced machine learning developers and researchers alike. By using Ludwig, experts and researchers can simplify the prototyping process and streamline data processing so that they can focus on developing deep learning architectures rather than wrangling data.

Ludwig introduces the notion of data-type-specific encoders and decoders, which results in a highly modularized and extensible architecture: each type of data supported (text, images, categories, and so on) has a specific preprocessing function.

In this talk, we’ll be looking at leveraging TensorFlow -Extended (TFX) pipelines to programmatically create Deep Learning models with Ludwig for different input data types for both model training and inference using Beam-Flink-Python SDK.


Suneel Marthi

Principal Technologist, AWS
Suneel is a Member of Apache Software Foundation. He's previously presented at Flink Forward Berlin, Hadoop Summit Europe, Berlin Buzzwords, Open Source Summit, Big Data Tech Warsaw, and Beam Summit in the past.

Wednesday October 30, 2019 11:30 - 12:05
Rhone 1


Explaining the Black Box of Machine Learning Models with Alibi - Janis Klaise, Seldon
Being able to reason about the predictions of a machine learning system is becoming increasingly important as sophisticated, non-linear predictive models are being adopted across the enterprise and beyond. In this talk we will discuss some requirements and challenges of model explanation algorithms and demo some practical examples using the open-source library Alibi we've developed at Seldon.

- What makes an explanation interpretable?
- The trade-off between interpretability and fidelity of an explanation algorithm
- Practical examples of using some interpretable techniques (e.g. anchors, counterfactual search) for classification of tabular data, text and images using the open-source library Alibi


Janis Klaise

Data Scientist, Seldon
Janis Klaise is a Data Scientist at Seldon primarily working on algorithms to provide rich information beyond raw predictions for live ML systems (e.g. model explanations, outlier detection, model confidence, concept drift). Ongoing projects include the development of the open-source... Read More →

Wednesday October 30, 2019 12:20 - 12:55
Rhone 1


Natural Language Processing with Deep Learning and TensorFlow - Barbara Fusinska, Google
Natural Language Processing offers a variety of techniques to get insight from and generate text data. Going beyond simple representations and taking advantage of Deep Learning and RNNs, the models can use document context to perform more accurately. With the help of libraries like TensorFlow, building neural networks and applying NLP is now available to the wider audience.
In this session, Barbara will make the introduction to NLP concepts and deep learning architectures. The audience will be walked through two labs: sentiment analysis and text generation.
After this session, the audience will have a good understanding of the deep learning concepts when it comes to NLP. The attendees will create a classifying model that takes advantage of the document context using TensorFlow library.

avatar for Barbara Fusinska

Barbara Fusinska

Strategic Cloud Engineering Manager, Google
Barbara is a Strategic Cloud Engineering Manager at Google with strong software development background. While working with a variety of different companies, she gained experience in building diverse software systems. This experience brought her focus to the Data Science and Machine... Read More →

Wednesday October 30, 2019 14:25 - 15:50
Rhone 1


The State of Production Machine Learning in 2019 - Alejandro Saucedo, The Institute for Ethical AI & Machine Learning
The machine learning ecosystem has been growing at break-neck speed making it harder to navigate. Similarly, as machine learning gets deployed in production, we find ourselves trying to address new complexities at higher states. It is critical that we idenfiy the different areas of focus for robust, reliable and scalable machine learning systems that require an intersection of best practices from the worlds of data science, devops and software engineering.

In this talk we dive into the state of production machine learning in 2019, where we start by defining the motivations for the talk, including challenges of ML at scale such as reproducibility, versioning, explainability, privacy, adversarial robustness and beyond. Then we proceed to dive into three of the most critical and broadly discussed areas of produciton ML in 2019, which include Model Orchestration, ML Versioning and Explainable AI. For each we define the ecosystem of open source tools and provide a hands coding example.

avatar for Alejandro Saucedo

Alejandro Saucedo

Engineering Director, Seldon
Alejandro is the Chief Scientist at the Institute for Ethical AI & Machine Learning, where he leads the development of industry standards on machine learning bias, adversarial attacks and differential privacy. Alejandro is also the Director of Machine Learning Engineering at Seldon... Read More →

Wednesday October 30, 2019 16:15 - 16:50
Rhone 1