Steve Pousty
Principal Community Architect, Red Hat
Actions
Steve is a Dad, Partner, Son, and Principal Community Architect focused on AI. He can teach you about Data Analysis, Computer Vision, AI, Kubernetes, Java, Python, and PostgreSQL. He has deep subject area expertise in GIS/Spatial, Statistics, and Ecology. Before Red Hat, Steve was a developer Advocate for Crunchy Data, DigitalGlobe, VMWare, LinkedIn, deCarta, and ESRI. Steve has a Ph.D. in Ecology and can easily be bribed with offers of bird watching or fly fishing.
Links
Kube and SQL - data science done right
One is the new hotness and the other might be seen as old and stuffy - yet both Kubernetes and RDBMS are two powerful tools to have in your data scientist toolbox (or to help give to the data scientists you support). PostgreSQL will be the particular database we discuss but most of the points apply to any modern RDMS. There have been a lot of overlooked features added to PostgreSQL that will bring joy to data scientists.
The goal of this talk is to help you appreciate all the ways that adding Kubernetes and PostgreSQL to your team’s tools can improve everyone’s lives by:
Less data over the wire
Doing the analysis where the data lives
Repeatability
Horizontal scaling for analysis (even autoscaling)
Most importantly, cleaner separation of concerns
By the end you will understand why both your data scientists, your ops people, and even your application developers will love bringing these two well known tools together to do a majority of their work. Come for the questions, feedback and fun banter!
Helping Data Scientists and Ops get along
There has been quite a bit of focus on devops and its promise of making life better for both app devs and ops people. There is a growing movement for dataops, trying to bring some devops repeatability, security, and agility to the data world. Most of this work has focused on the pipeline of going from initial data to presentation to line of business owners.
This talk is going to focus on something a little different. We are going to look at how SaaS, APIs, and automated workflows can help manage scenarios for deployments to production. The goal of this talk is to show a possible cloud native architecture that reduces friction between the data team and the ops team. Our specific use case will cover allowing the data team to safely publish their models on edge machines, while allowing the ops people the visibility and safety they love.
If you want to see a fun example of infrastructure code leading to world peace then come on by. There will be live running code, interesting life lessons, and questions will be encouraged.
Building cloud native spatially aware agent based simulations
Agent-based simulations are used in a wide variety of fields, from biology to economics. An agent-based simulation gives individual actors rules at the actor level but then we look for emergent patterns and behavior at a macro level . Frequently these actors can move in 2D and 3D space, so these models tend to have a strong spatial component
Over the last 30 years we have seen the rise of several different spatially aware agent-based simulation frameworks, such as Mesa (Python) or Mason/GeoMason (Java). During that same time we have seen the emergence of vertically and horizontally scaling of distributed applications. There have also been many advances in rapidly handling large streams of data as well as loosely coupling pieces of an application.
The talk will cover a brief introduction to spatially aware agent-based modeling. The rest of the talk will cover building an agent based simulation for finding optimal animal corridor locations. We will cover the architecture, then show the agent running, and then visualize the results. The application will take advantage of data streaming, messaging, and horizontal scaling to create a robust and efficient framework.
The goal of this talk is for you to leave understanding spatially aware agent-based simulations and how they can leverage modern cloud capabilities. It doesn’t matter if you are in finance, social media, or pharmaceuticals - agent-based modeling can help your understanding of your vertical. With this knowledge in hand you will sure to be a hit at your next cocktail party. Come for the knowledge and stay for the witty banter.
Composing Modern Realtime Spatial Applications
Doesn’t matter what you call these applications: Command and Control, Situational Awareness, Real Time Tracking - they are super duper cool and probably the surest mapping apps to impress friends and neighbors (and bosses and clients).
There is a bunch of recent tech that makes creating these types of apps easier. We are going to use “revenge of the boring” tech, PostGIS and Java, together with some “new hotness”: pg_tileserve, pg_featureserve, pg_eventserver, and Kubernetes, to create a sweet realtime dashboard. Our application will track buses from the Snohomish County Public Transportation Benefit Area with live updating locations.
The talk will give a brief introduction to the data, discuss the architectural choices, show you the application, and then dig in on how we did it. You know you have always wanted to build one of these types of applications. Whether it’s buses, safety personnel, or your cat, come and learn how easy it has become for you to create web-based tracking applications. Questions and discussions will be encouraged
Bringing AI/ML to your Enterprise App in the Cloud
You’ve seen the cool demos using machine learning. You see everybody talking about how amazing it is and how it’s going to change the world. And there you are, building your money making, salt-of-the-earth, Java enterprise application. When do you get to do the cool stuff?!
The time is now! This talk is going to show you some of the ways you can simply leverage cloud based machine learning and AI to add pizzazz to your apps and a bounce in your step. While we will be using Azure services in this talk - all the major hyperscalers can give you this ease of use on their platforms.
We will cover two different use cases. The first will be tagging and sorting a large catalog of images. The second will be some sentiment analysis on messages from Twitter Mastodon.
We will take some relatively simple Spring code and then bake in Machine Learning functionality. The big win here is that you don’t need a team of data scientists, large sets of training data, specialized authentication, or a dataops team building your own custom pipelines. This functionality will come out of the box ready to go!
The goal of the talk is to get you familiar with how easy it has become to add machine learning to your everyday applications. We will start with a brief introduction to Machine Learning terms and techniques, then give our use case, data, and code, and then you will see it all running in action.
Come for the information but I hope you stay for the discussion
Azure Spring Apps Enterprise - all the Spring Goodness and none of the management headache
You love Spring (or .NET or Node) but you hate managing cloud infrastructure. You know you should run your stuff on one of the hyperscalers, but you don’t want to figure out how to put together their vast offerings into something useful. Really, you want to put your Spring app “in the cloud”, take advantage of some of the services they offer, such as machine learning or messaging, and not have to deal with operational issues.
If this is you, then this is the talk for you. We are going to take a simple Spring Boot application and run it in Azure Spring Apps Enterprise (ASA-E). Getting the app on the platform is the easy part, the most fun parts of the demo will be how to integrate our application with capabilities from the ASA-E and the Azure platform. We will look at using hosted Spring features such as Tanzu Spring Cloud Gateway as well as native Azure features, such as Azure Service Bus.
Or maybe you are a DevOps engineer in your company and a lot of your users are giving you Spring apps they want to take to production. There will be numerous places I will demonstrate or discuss the value to the devops team of using ASA-E to manage and run their application.
Along the way there will be lively questions and answers. The goal of this talk is to set you up so you can go forth and conquer after the show. You will dazzle and amaze your co-workers with your ability to go from Spring Boot application to Cloud Native Application without ever touching Kubernetes.
Becoming a good speaker
There is plenty of material on how to give a good talk. As a developer advocate, your relationship with public speaking will likely grow beyond giving a good talk to becoming a good speaker. We are going to focus on how to grow your craft.
We will start with examining some ideas about your audience and what they want. Then we will cover some of my thoughts on giving a good presentation. From there we will explore some of the work you need to do in becoming a “good speaker.”
While you won’t be a great speaker by the end of the talk, the goal is to give you some things to think about and some practices you can use to make better talks and become recognized for the craft of public speaking.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top