When building reliable services unexpected user behavior can prove deadly. Yet few engineering teams make use of talent from their product and design orgs if they are working on something internal. This talk will demonstrate how a product mindset and design thinking can super charge technical discussions and build stronger systems. Don't rely on the user to be "smart!" It can trap you in an impossible architecture.

I'm a software engineer who spends her time writing code and developing apps. I have a pretty good grasp of the vocabulary and technologies relevant to my job. But what happens when another facet of engineering, one that is gaining a lot of traction and has a large footprint of its own starts becoming more and more relevant to my day to day tasks? Well, that's exactly what happened to me last year when DevOps became a big part of my role. This is my attempt to impart all of that knowledge onto you.

Taking advantage of AWS Parameter Store and some automation magic to make secret change management a problem of the past. We will talk about:

• The importance of secrets management
• Use Cases
• Parameter Store vs. Secret Manager
• Secret Vault and Application Integration points
• OSS Tools for Automation

Resilience Engineering (RE) is both multidisciplinary field of study as well as a community of practitioner-researchers from multiple high-tempo, high-consequence domains such as aviation, medicine, power distribution, space operations, and critical infrastructure. In recent years, the world of software engineering and operations has become involved and more acquainted with this almost 20 year-old field and community. This talk will give a “lay of the land” on what RE is, what it is not, what efforts are relevant to the DevOps community, what attendees can do to learn more.

As developers, we take great pains to make our applications accessible to users on a range of devices and browsers. There are more adults with disabilities in the US than there are users of Internet Explorer. So why do we spend more time worrying about polyfills for out-dated browsers than we do on ensuring the accessibility of our applications for users of all ability levels? This talk will emphasize the importance of incorporating accessibility best practices throughout our design and development workflow. Simulated experiences of engaging with inaccessible software will provide insight into the realities that millions of users face day to day. Drawing on recommendations from the Department of Justice, we will discuss action steps for evaluating and improving the accessibility of software.

Kim is an attorney and disability rights advocate. Her interest in accessible technology stems from her own vision impairment and her use of assistive software on a daily basis. In her career she has worked in various disability policy areas including housing rights and access to reasonable accommodations in employment. Kim has dedicated her career to the full inclusion of people with all types of disabilities.

Today--Incorporating AI into applications is as easy as a single API call. Once imbued with AI, these transformed applications can improve over time as they learn from user interactions. How do we manage these new learning systems to ensure that they take advantage of all available information while maintaining accuracy and minimizing bias? This talk will discuss common problems encountered when designing and implementing AI systems and how DevOps practices can be used to address these issues. DevOps practitioners will learn the relevance of their skillset and practices to the rapidly evolving AI domain. The problem of properly managing AI applications and their underlying models is not trivial.

Michael Flores is an Architect on the Strategy and Technology Team in IBM’s Federal Chief Technology Office. He has helped numerous agencies use Cloud, Artificial Intelligence(AI), Robotics, and DevOps through the adoption of open standards such as Cloud Foundry, Docker, and OpenWhisk. Michael represents IBM in The Open Group as chair of the Open Platform 3.0™ Forum, where he collaborates with other industry leaders to develop standards for modern technologies to support global interoperability and boundaryless information flow. Michael has extensive development, design, and architecture experience across a variety of technologies and has implemented numerous chatbots and AI driven solutions through various channels including web, SMS, and robots.

Michael received an undergraduate degree in Bioinformatics from Baylor University and holds a patent for dynamic customization of reference architectures. His capstone project at Baylor focused on crawling heterogenous data sources to create knowledge graphs to aid PTSD research.

Working technology for a political campaign involves the shortest timelines, tightest deadlines, and highest stakes you will likely ever encounter in a technology career. Come hear a tale of two political campaigns - a state measure campaign and a presidential campaign - and the application of both DevOps technologies and culture to move fast, pivot quickly, and hopefully win. One of the key challenges of politics - as well as DevOps in general - is harnessing automation without losing the critical human touch which moves hearts and changes minds. Learn how to find the line where too much automation (yes, there is such a thing) is counterproductive and you need to pull back to maintain a personal connection with voters, customers, employees, and more. You will also walk away knowing how to take the lessons and experience learned to future campaigns and projects - especially when your candidate, product, etc. does not end up winning. There is value - sometimes more value - in a loss as well as a win. Learn how to take what you can, iterate, and refine it for a future application.

GDPR is upon us, along with new requirements for protecting user data. Many of us in the DC ecosystem are no strangers to compliance! We have expertise in HIPAA, SOC, PCI, and of course the stringencies associated with government work. But GDPR's new requirements are taking effect at the same time that many users are demanding increased privacy and transparency about their data (as a backlash to the Cambridge Analytica scandal) regardless of regulation. We'll look at best practices for DevOps teams to respond to this increased need for privacy, transparency and security around user data.

In addition to best practices, we'll explore the following questions:

What does the current regulatory landscape look like, and how does that impact DevOps?

How do we design our applications and our architecture (especially containers) to optimize for security and privacy?

And are the current regulations helpful as guidelines, or will they become another set of checkboxes?