Steven Pousty

Youve seen the cool demos using machine learning. You see everybody talking about how amazing it is and how its 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 dont 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

In this workshop, we are going to get hands on and build a Dungeon Master (5th Edition) built with a vector database to help us carry out Retrieval Augmented Generation (RAG).

We'll run all this on Kubernetes. We will also be using Langchain4j. Let's see how well “AI” can handle being your dungeon master.

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.

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 teams tools can improve everyones 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!

In the last two years, AI machine learning has exploded in prominence. One of the key concepts used in the modeling and storage of AI is vectors. There is no doubt vector data management will be a key concern of most people concerned with the lifecycle of data management. Feeling like you should learn more and how you would use them in your data work? Then have I got a talk for you!

We will start by explaining the concept of (embedding) vectors and how they are used in the AI life cycle. From there we will go into putting them into a database.

Now that those vectors are in a database we can talk about the use cases where the technology makes sense. As opposed to an RDBMS, vector databases are more tightly focused and optimized for particular use cases.

To ground this discussion in something more concrete, we will show live demos of the technology throughout the talks. By the time you leave you will have a strong base to go home and explore more (and impress friends at dinner parties).

You know, and probably already love, PostgreSQL as your relational database. But this session will show you all the other features you never knew Postgresql brought to the table. We will show you how you can forget about using ElasticSearch, MongoDB, and Redis for a broad array of use cases. The end goal of the session is to show you how Postgresql should be 80 of the 80/20 rule when choosing a datastore. Unless you have a specialized use case, PostgreSQL is the answer.