Agentic Conversations (formally mlops.community)

Running a Fintech on Machine Learning

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter⁠

For this meetup, we sat down with Caique Lima and Cristiano Breuel, Machine Learning Engineers at the Brasilian Fintech Nubank.   

Nubank is a Fintech providing credit and banking services to more than 20 million customers. Data science has been one of the company's pillars since the beginning, and many of its critical decisions in production are made with ML, in areas such as Credit, Fraud, and Customer Service. We discussed how they develop, deploy, and monitor ML models, and also talked about how they built those in-house solutions over the years. Today, they use MLOps to support a team of more than 70 Data Scientists/ Machine Learning Engineers.  

Caique is a Machine Learning Engineer at Nubank, developing software to scale decision-making, which goes from model development to monitoring. Always trying to bring good practices from Software development to Data teams.  Cristiano, a Machine Learning Engineer at Nubank, works to improve the efficiency and quality of ML development. Previously ML/Data Engineer at Google, specializing in MLOps, and Software Engineer at IBM.  

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Cris Sterry on LinkedIn: https://www.linkedin.com/in/chrissterry/
Connect with Cristiano on LinkedIn:  https://www.linkedin.com/in/cristiano-breuel/
Connect with Caique on LinkedIn: https://www.linkedin.com/in/caiquelima/

DataOps and Data Version Control  

MLOps.community meetup #19 with the Founder and creator of DVC.org Dmitry Petrov. 

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter⁠

Data versioning and data management are core components of MLOps and any end-to-end AI platform. What challenges are related to data versioning, and how to overcome? What are the benefits of using Git and data codification as a foundation of data versioning? And how open data versioning tools can enable an open MLOps ecosystem instead of closed end-to-end ML platforms.  

DVC and other tools:
Basic modeling scenarios, Automation of modeling, Model deployments: to server or Docker.
DVC as a model registry.
CI/CD for ML  
Dmitry is a creator of the open-source tool Data Version Control - DVC.org - or Git for data. He is a former data scientist at Microsoft with a Ph.D. in Computer Science. Now Dmitry is working on tools for machine learning and data versioning as a Co-Founder and CEO of Iterative.AI in San Francisco, CA.  

Join our Slack community: https://go.mlops.community/slack

Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Dmitry on LinkedIn: https://www.linkedin.com/in/dmitryleopetrov/

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

MLOps coffee sessions coming at you with our primer episode talking bout KFserving!   David Aponte and Demetrios Brinkmann dive deep into what model serving is in machine learning, what different types of serving there are, what serverless means, API endpoints, streaming and batch data, and a bit of coffee vs tea banter.

||Show Notes||  

ML in Production is Hard Blog article by Nikki: http://veekaybee.github.io/2020/06/09/ml-in-prod/?utm_campaign=Data_Elixir&utm_source=Data_Elixir_289
Interactive learning platform Katacoda: https://www.katacoda.com/
Github repo used in video: https://github.com/aponte411/demos
Blog on different ways to handle model serving: http://bugra.github.io/posts/2020/5/25/how-to-serve-model/  

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with David on LinkedIn: https://www.linkedin.com/in/aponteanalytics/

MLOps.community meetup #17: a deep dive into the open source ML framework Hermoine built on top of MLflow with Neylson Crepalde  

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

Key takeaways for attendees:

MLOps problems are dealt with tools but also with processes. Open-source framework Hermione can help in a lot of parts of the operations process  

// Abstract:
In Neylson's experience with Machine Learning projects, he has encountered a series of challenges regarding agile processes to build and deploy ML models in a professional cooperative environment that fosters teamwork. While on this journey, Neylson and his team developed some of their own solutions for these challenges.  Out of this was the open-source project Hermoine born. Hermoine is a collection of solutions for these specific MLOps problems that were packaged into a library, an ML project structure framework called Hermione.  

In this meetup, we talk about these challenges, what they did to overcome them, and how Hermione helped address these different issues along the way. We will also do a demo on how to build an ML project with Hermione.  
Check out Hermoine here: https://github.com/a3data/hermione  

Neylson Crepalde is a partner and MLOps Tech Lead at A3Data. He holds a PhD in Economic Sociology, a Master's in Sociology of Culture, an MBA in Cultural Management, and a bachelor's degree in Music/Conducting. He is a professor of Machine Learning and Head of the Data Science Department at Izabela Hendrix Methodist Technological University. His main research interests are Machine Learning processes, Politics and Deliberation, Economic Sociology, and Sociology of Education. In his PhD, he has worked with Multilevel Social Network Analysis and Exponential Random Graph Models to understand the social construction of quality in an orchestra’s market.  

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Neyslon on LinkedIn: https://www.linkedin.com/in/neylsoncrepalde/

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

Venture Capital in Machine Learning Startups With John Spindler, CEO of Capital Enterprise. 

John Spindler, CEO of Capital Enterprise. We talked about what trends he has been seeing within MLOps, ML companies, and also how he evaluates a deal. 

John Spindler has over 15 years of experience as an entrepreneur and business advisor/consultant, and, as well as being responsible for the day-to-day management of Capital Enterprise, he is also a general partner at AI Seed, an early-stage fund that invests in highly talented AI-first companies. 

John is on a mission to make it possible for someone moderately intelligent, with a good idea, ambition, and passion, to make it as an entrepreneur. 

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g   

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with John on LinkedIn:  https://www.linkedin.com/in/johnbspindler/

Human In The Loop Machine Learning and how to scale it with Robert Munro. 

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

This conversation centered around the components of Human-in-the-Loop Machine Learning systems and the challenges when scaling them. Most machine learning applications learn from human examples. For example, autonomous vehicles know what a pedestrian looks like because people have spent 1000s of hours labeling “pedestrians” in videos; your smart device understands you because people have spent 1000s of hours labeling the intent of speech recordings; and machine translation services work because they are trained on 1000s of sentences that have been manually translated between languages.

If you have a machine learning system that is learning from human feedback in real-time, then there are many components to support and scale, from the machine learning models to the human interfaces and the processes for quality control.   

Robert Munro is an expert in combining Human and Machine Intelligence, working with Machine Learning approaches to Text, Speech, Image, and Video Processing. Robert has founded several AI companies, building some of the top teams in Artificial Intelligence. He has worked in many diverse environments, from Sierra Leone, Haiti, and the Amazon, to London, Sydney, and Silicon Valley, in organizations ranging from startups to the United Nations. He has shipped Machine Learning Products at startups and at/with Amazon, Google, IBM & Microsoft.  

Robert has published more than 50 papers on Artificial Intelligence and is a regular speaker about technology in an increasingly connected world. He has a PhD from Stanford University. Robert is the author of Human-in-the-Loop Machine Learning (Manning Publications, 2020)  

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g   

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Robert on LinkedIn: https://www.linkedin.com/in/robertjmunro/  

Robert's book on Human in the Loop Machine Learning: https://www.manning.com/books/human-in-the-loop-machine-learning
Blog Post "Active Learning with Pytorch": https://medium.com/pytorch/https-medium-com-robert-munro-active-learning-with-pytorch-2f3ee8ebec

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

The amazing Byron Allen talks to us about why MLflow and Kubeflow are not playing the same game!
ML flow vs Kubeflow is more like comparing apples to oranges, or as he likes to make the analogy, they are both cheese, but one is an all-rounder and the other a high-class delicacy. This can be quite deceiving when analyzing the two. We do a deep dive into the functionalities of both and the pros/cons they have to offer.

Byron is a Senior Consultant at Servian - a data consultancy in Australia that also has a footprint across APAC as well as the UK. Byron is based in the London office, where he helps organizations discover and build competitive advantage through their data. His focus is on client advisory and consulting delivery related to Experiments and ProductionML (i.e., data science, experimental design, ML model development, MLOps).
Byron has written about a wide range of topics, including the divide between data engineer and scientist, the role of ML in the post-COVID world, and Kubeflow vs. MLflow. Check it all out here: https://medium.com/@byron.allen

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Byron on LinkedIn: https://www.linkedin.com/in/byronaallen/

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

Resume building and Interviewing tips for data scientists and Machine learning engineers.  

When on the job hunt, there are some tested tips and tricks that can be applied to your resume and interviews, which will give you a leg up on the rest of the competition.   

Anthony Kelly, host of the AI in Action podcast and Executive Search Consultant focused on Machine Learning and Data Science, sat down with us to talk about what some of the best resumes and CVs have in common. 

We spoke about optimizing your CV/resume and maximizing opportunities once you land an interview, so you can have the most amount of options to choose from while you are on the market.  

Anthony Kelly is a tech recruiter from Dublin, Ireland.  Currently, he is the Country Manager for Alldus, an international AI and Data Science Recruitment company. He has been working in recruitment since March 2015, and since joining the recruitment industry, he has been an international top recruiter due to his performance.

Along with the above, he is also the founder of the Berlin AI in Action meetup, which has over 2,000 members. In addition to the meetup, he also has a podcast series called AI in Action and is a Co-Founder of the Berlin AI community Awards.

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Anthony Kelly on LinkedIn: https://www.linkedin.com/in/anthonypierrekelly/  
Check out the AI in Action podcast: https://podtail.com/podcast/ai-in-action-podcast/

MLOps meetup #12 // What are the advantages for a data scientist to know data engineering?  

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

What good is learning Data Engineering skills? These days full full-stack is overflowing with all the different things you need to know about, so why learn data Engineering now? Our guest on this meetup will make the case for what the advantages are if you do decide to learn data engineering, and also go into depth on how to do data engineering in the cloud.  

Dan Sullivan is a software architect and data scientist with extensive experience in big data, machine learning, data architecture, security, stream processing, and cloud architecture. Dan is the author of the official Google Cloud study guides for the Professional Architect, Professional Data Engineer, and Associate Cloud Engineer exam guides, as well as NoSQL for Mere Mortals.  

He is also the author of over ten LinkedIn Learning courses on data science, machine learning, SQL, data architecture, and NoSQL. He holds a Ph.D. in genetics, bioinformatics, and computational biology.  

Get a copy of his new book here: https://www.wiley.com/en-us/Official+Google+Cloud+Certified+Professional+Data+Engineer+Study+Guide-p-9781119618454  

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Dan Sullivan on LinkedIn: https://www.linkedin.com/in/dansullivanpdx/

MLOps community meetup #11 Machine Learning at scale in Mercado Libre with Carlos de la Torre.

Join the Community: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTJoinIn⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Get the newsletter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://go.mlops.community/YTNewsletter

Mercado Libre hosts the largest online commerce and payments ecosystem in Latin America. The IT department built Fury: a PaaS framework for the development and deployment of multi-cloud, multi-technology, microservices. This platform leveraged the growth of the IT area, which now counts ~4000 people.   

As such, it lacked support for machine-learning based solutions: an experimentation environment for data scientists, infrastructure and data access support for ETL and models’ training tasks, etc.  Therefore, for over a year now, they have been developing Fury Data Apps (FDA). An extension of Fury for the design, experimentation, development, and deployment of machine-learning-based solutions. It is already supporting ~500 users and some high-performance production APIs.  

In this meetup, we talk about the main features of the platform, the supporting technology, and why Carlos never accepted my LinkedIn request.
Link to the article Carlos references: https://martinfowler.com/articles/cd4ml.html 

Join our Slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://zoom.us/webinar/register/WN_a_nuYR1xT86TGIB2wp9B1g  

Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Carlos on LinkedIn: https://www.linkedin.com/in/carlosdelatorre/
Follow Carlos on Twitter: @py_litox