DevNation Federal: Data Science Unleashed

In this session, we will discuss real, tangible use cases for leveraging artificial intelligence and applied machine learning models to solve problems that affect everyone. We will walk through what it takes to build software that serves more people by creating inclusive teams that are empathetic and courageous.

Software engineering has changed dramatically over the years with better collaborative development practices, migration from monoliths to microservices and more. But what about machine learning? How has machine learning based development evolved over the years? In this session, we'll discuss early ML setup and practices and how it's evolved into today's collaborative data science and streamlined ML deployment and MLOps best practices.

As opportunities for data science and machine learning flourish across the enterprise, there’s a growing friction between desires to build new systems and responsibilities to extract more value from legacy systems. With a survey of experiences across industries over the last decade – including Federal Government, Oil & Gas, Financial Services, and Precision Healthcare – we’ll explore what it takes to succeed with a modular open systems approach and how it can ease the friction.

Quantum computers today are expensive, rare, highly specialized resources, but hold tremendous promise for the future. Realizing this promise will require deep consideration of how to enable useful access to these resources, how to enable all developers (not just quantum specialists) to use them, and bridging classical and quantum applications for developers and operators alike. We’ll discuss the constraints QCs have and how building a hybrid, DevOps approach can unlock quantum computing in the future.

Companies everywhere are implementing DevOps practices to bring fast innovation of software to market. Frequent pushes of small, microservice updates are in high demand. Federal agencies want in on the action too, and they bring their own sets of challenges.

When working within air-gapped environments, security is of the utmost importance, so how do you make it work? Open source solutions don’t always seem like the first choice to turn to. Many of them demand constant Internet access for updates. With the right strategy in place, open source solution can often not only be an option, but the best option. In this session, Jay Hancock, DevOps Manager at Parsons Corporation, will discuss how he and his team have addressed some of the challenging requirements of the defense and intelligence community. He will provide a detailed viewpoint into using open source technologies like kubernetes, containers, and more.

Realizing when mechanical devices (e.g. pump data, manufacturing devices) are having mechanical issues and are showing signs of failure (via sensors) are important items to know when you need to keep them running 24/7. Therefore, creating failure prediction algorithms and/or model(s) are essential parts of your mechanical equipment maintenance toolbox, but how do you get started?

Failure prediction in real time on time series data (gathered by sensors) can be realized with the use of Open Source tools. In this session we will take you through creating data science applications on the edge and how we secure data, models and edge devices.