ReDev is Boston’s premiere conference for Developers and Technical Executives who are striving to push limits and bring their organizations to the next level.
ReDev brings together the entire product chain workforce, from senior level strategists to secure implementation experts and product leaders to the development teams that create the impossible from keystrokes.
Federal Reserve Boston 600 Atlantic Ave Boston, MA
May 16th, 2018 9:00am-5:30pm
Lunch Keynote Panel
Building & Scaling Teams @ Amazon
When Slides or Videos are available, they will be linked in the expanded descriptions.
- +Variable Fonts and Our New Typography
For centuries, typography has shaped the way we ‘hear’ what we read. In our web work, though, we’ve have to balance our typographic desires with user experience and performance, knowing that every weight, width, or style of a typeface required a different file download. Variable fonts change that, as they include _every_ width, weight, slant, and other permutation of a typeface, all in a single file not much bigger than a regular font file. Now, beautiful web typography can be crafted to respond to screen size, language setting, even ambient light. In a detail-packed talk, Jason will show you not just how far the new capabilities can take us, but how to make use of them right away.
- +Old Company, New Technology: Using Elixir at the MBTA
Gene Shkolnik – MBTA
- +Does your DevOps feel a lot like Office Space?
In this session, we will review the demands on Developers, in an organization that is working through an Agile/DevOps transformation. For Developers, trying to develop solutions, while meeting the demands of the company, can be a significant challenge. This session will showcase how to ensure that Developers can get code running in Production as quickly as possible while meeting all the requirements of the organization. We’ll make sure you know how to keep the Lumbergs off your back.
TJ Randall – XebiaLabs
Erik Rasmussen – XebiaLabs
- +The Evolution of Development and Delivery at Circle
Circle’s environment consists of an ever growing number of microservices. Developing and testing new services while building and releasing changes for existing services as they scale has been an interesting challenge. In this session we’ll discuss how they’ve engineered their development environment for rapid releases and empowered teams to collaborate on service development.
Kevin Karwaski – Circle
- +From reality to mixed reality - A developer’s journey
Learn the oohs and aahs of creating AR/VR apps from the perspective of a (web) engineer. Get an inside look into how an online retailer went from 2D to 3D, and is creating the next generation of shopping experiences. We’ll discuss the process, the tools, the challenges, and the rewards!
Shrenik Sadalgi – Wayfair
- +Developing for the Healthcare Space
When building products for the healthcare space, having a bug can mean losing FDA approval or even losing a life. While engineers can often use their skills to build any number of tools when they jump company to company, what extra training is needed for healthcare and how much time should be dedicated to building up clinical knowledge of the problem you’re solving or working with academic experts. How can you push for innovation in a highly regulated space?
This session will pull together a panel of experts who are building the products and the people who are charged to implementing them into hospitals.
Jesse Johnson – Verily
Mark Nesline – Imprivata
Sarah Sossong – Mass General Hospital/Partners HealthCare
Roy Schoenberg – American Well
Cory Parent – Cota Healthcare
- +Edge Computing: Thinking Beyond the Data Center
For over a decade, centralized cloud computing has been long been considered a standard IT delivery platform. Recently companies have been applying the simplified administration and flexibility of cloud computing architectures to distributed infrastructures that span across multiple sites and networks. Edge Computing is emerging as a new way of delivering IT Infrastructure for use cases that were never imagined. Think about the possibility of a global swarm of hundreds, thousands or even millions of nodes. Edge Computing requires radically different approaches and tools.
Use cases include:
• Extreme edge conditions (IoT, transportation, manufacturing, utility grids, VDI),
• Long distances between nodes (Satellites, Supply-chain, global shipping)
• Significant hardware and network connectivity constraints
• Hub and spoke designs (Telecom networks, global application distributions)
• We need to think differently about an Edge Computing ecosystem.
• Different architectures to different use cases: hierarchical, matrixes, widely distributed, or mesh.
• How node size affects performance and fit for each use case.
• Deployment considerations
• More stringent security (code/design) in exposed environments
Beth Cohen, Product Manager, Verizon
- +Ground Truth Data Problems in Business
Ground truth for data is frequently missing in business. It is either
unavailable, expensive to get, or private. This talk focuses on the
ubiquity of this problem in today’s world in which more and more of the data
production flows require Machine Learning/Artificial Intelligence
algorithms. How can we measure the accuracy of our algorithms/robots
when curated data is scarce or missing? We offer a universal
suggestion to solving these problems – bypass the need for detailed
knowledge of the ground truth for your data by estimating directly the
statistics of interest for research and business development. We
suggest that a future with smart robots will require that they measure
their own errors so they can function autonomously from humans.
We illustrate this approach with four real-world examples
- Focusing multiple aerial maps into a precise final map – ground truthunknown.
- Studying the dynamics of data flows in a large ad-tech database withHyperLogLog – ground truth expensive.
- Measuring the accuracy of a unique web ID service (super cookie)without user identity – ground truth private.
- Measuring the accuracy of binary classifiers – ground truth unknown.
Andres Corrada-Emmanuel, Data Engines Corporation
- +Bots: Virtual and Intelligent Agents for Better Self-service
Virtual and Intelligent Agents a.k.a. Chatbots are becoming ubiquitous in our daily lives. Alexa can turn on your favorite music at the time that you want, and reserve dinner at your favorite restaurant. Life simplified and personalized is the new frontier of digital innovation. So it is no surprise, that a recent Becker’s Hospital Review survey shows that patients as consumers, want their digital healthcare experience to be more accessible and personalized like retail. Included in the six areas consumers said improvement is most needed are, searching for a physician or specialist (81%) and changing or making an appointment (79%), both very critical to improving patient access management.
At Kyruus, we are integrating Virtual Agents into our solutions to improve patient outcomes by matching patients to the right providers through intelligent, automated patient routing and self-service appointment booking capabilities while delivering a highly personalized and engaging patient experience. Adoption of Virtual Agents allows health systems to not only boost patient acquisition, retention, and conversion but also to operate more efficiently by focussing their human agents and providers on high value patient engagements.
In this session, you’ll learn what’s feasible and pragmatic in leveraging Chatbots –
● Chatbots as a distinct UI channel and insights into how to use them to best serve users
● Chatbot framework and selection criteria for type of Chatbot to use
● Conversation design considerations and success factors
● Technology choices and representative architecture for Chatbot integration
● Challenges and lessons learned
● Implications specific to Healthcare
Nausheen Moulana, Kyruus
- +The UX, PM & Engineering Triad
Building great products is a cross functional effort. Product formulation, product design and product delivery all go hand in hand in making great products. The collaboration between these three groups is critical to the success of any product. Often times these three teams operate in silo’s – but need to work together to create greatness. This panel will explore this approach, how to make it work, how to avoid pitfalls and creating a culture of collaboration.
Lee Weiner, Rapid7
Colleen Miller – Toast
Pryia Sapra – Shyft Analytics
Margaret Olsen – Cogito Corp
- +# Free and open source in 2018
From its creation in 1983, to the creation of the Linux kernel in the early 1990s, free software moved at an incredible pace. While the advent of the web and the ubiquity of high-speed of high speed Internet access has made collaboration easier than ever, has the free software dream failed to realize? Matt Lee examines the brief 35 year history of the free software movement, from GNU and open source and beyond.
- +Managing Project Uncertainties with Quantitative Risk Analysis
Increasing level of uncertainties affects technical projects on daily basis. Market changes can add new requirements, technological complexities can create a new level of “unknowns”, global inter-dependencies and distributed teams can bring communication challenges, organizational changes can cut project resources… These are just a few factors that can add risks throughout project execution and derail project success.
• How these negative events can affect your project objectives?
• What is the confidence factor that you will deliver on time and within the budget?
• Which risks should be addressed first?
• What is the effectiveness of your mitigation strategy?
The answers to these questions are very important for choosing effective response strategies and keeping your project on track. This presentation will show how Quantitative Risk Analysis can help better manage project risks in today’s uncertain world.
Target audience for this presentation includes technical project/program managers as well as engineering managers leading technical projects.
Yelena Okonechnikova, Qualcomm
- +Forget about AGI, let's make usable AI tools!
Due to some great successes and mounting excitement in the field of Machine Learning and Artificial Intelligence research, there’s a lot of discussion these days about whether human-level AI, or Artificial General Intelligence (AGI), is achievable in the coming decades. There are podcasts devoted to exploring how it might come about, Elon Musk is expecting it within a single decade. But some, including AI researchers, are wondering whether the current excitement over AI constitutes a bubble, and if so whether it’s about to burst.
If there are expectations that conversational interfaces like Alexa and Google Home Assistant are going to reach human levels of intelligence in anything like the near term, these expectations will surely remain unmet. Failure to meet expectations is what brought on so-called “AI winters” in the past. But although Alexa and Google Home Assistant will likely continue for a long time to be as dumb as they are today, this doesn’t mean they can’t be made a lot more useful. This talk will look at ways in which vast improvements can be made in Machine Learning applications without significant progress being made towards human-level intelligence or AGI.
As Machine Learning techniques are incorporated into more and more applications, being used by more and more people, the unique challenges they present will become increasingly important to tackle. One such challenge is the problem of uncertainty. The uncertainty inherent in the predictions of ML systems is not going to go away, and so it must be addressed in any app that uses such predictions. While users of ML systems would rather have clear-cut answers to their questions than estimates or confidence intervals, misrepresenting ML answers as definitive can be highly problematic. Another big challenge is the problem of explainability, or interpretability, of ML predictions. There are numerous approaches being tried but some raise the question of whether the purpose of an explanation is just to reassure, rather than actually provide an accurate insight into why a system responded as it did.
Another area that is ripe for deeper exploration is few-shot learning with a human-in-the-loop. ML practitioners, working in collaboration with software engineers and UX professionals, can attain great successes by opening up Machine Learning to companies that don’t have vast quantities of labeled training data.
Katherine Bailey, Acquia
- +Creating Personal and Engaging User Experiences using AI and Cognitive Computing
Artificial Intelligence is permeating everything we do today from online shopping to medical research and beyond. During this engaging and immersive session we’ll explore how to use cognitive computing and AI to deliver truly unique and engaging user experiences. Capabilities such as facial recognition, sentiment analysis, speech synthesis, linguistic understanding and many more are key elements to breaking through and delivering incredible experiences to your users. This session will be exploratory as well as practical to insure attendees are empowered to start their journey on new human-machine experiences that will wow their users.
Tim Baggs, Microsoft
- +Data Protection with OpenStack and Containers
“We Don’t Need No Duplication” would be the lyrics, if we asked Roger Waters to describe OpenStack’s original design principles.
Planning for cloud native, ephemeral workloads, we paid less attention as a community to data protection. Fast forward to 2018, businesses, academia and service providers worldwide are deploying OpenStack for mixed workloads, part stateless and part stateful applications, making data protection a critical component.
- How severe is this problem?
- What comprises a good solution?
- How is the industry handling it?
- Is the Container community heading down the same path?
This 45-minute panel, featuring senior speakers from Red Hat, Verizon, Northeastern University and Trilio Data will provide a unique insight into this topic. We will learn from their experiences with First Class OpenStack deployments and hear about their vision for leveraging OpenStack and Container technologies to serve both persistent and non-persistent workloads. We will leave time for audience questions, take this opportunity to find out more!
Boaz Michaely – Trilio Data (Moderator)
Sean Cohen – Red Hat
Amrith Kumar – Verizon
Peter Desnoyers – Northeastern University
- +A Practical Guide to Creating More Desirable Products Through Collaborative User Research
Product development is a lofty topic with many branches to discuss, worthy of an entire day. For this session we’re focusing on User Research and how to find critical issues before they’re launched into production. To do this Mike will discuss:
- Introducing research methodologies to foster a more collaborative relationship with product managers, designers, clients, or other stakeholders
- Methods for catching critical usability issues before shipping your work
- Tools that can help facilitate usability testing and the discovery of post-launch defects once your product experience is released
- +Securely Testing, Deploying and Monitoring Software
How do you test, deploy and monitor software in a secure manner? Building security into our systems and software is a difficult undertaking which requires overcoming various hurdles along the way. In this panel, we will discuss common difficulties for building a security program such as: agile security, securing your tool chain, defense programming, static code reviews, testing (security & QA), solution deployment, monitoring, and auditing/reporting. Join us as we examine the pros and cons for solutions to these topics.
Neil Weitzel – Cygilant (Moderator)
Josue Diaz – IBM
Eitan Worcel – IBM
David Hazar – Deloitte
Umair Aziz – Creative Chaos
- +Achieving High-Performance, Mission-Critical Data Infrastructure Without a DBA
DraftKings is the premier daily fantasy sports provider. Servicing the needs of our customers, industry regulators and internal stakeholders comes with several unique technical challenges. A robust database layer is key to meeting those challenges. And how do we grow and maintain this critical infrastructure? Certainly not with a large team of highly experienced database engineers and administrators. In reality, DraftKings Engineering does not even have a single dedicated DBA on the team. We leverage the best RDBMS solution Amazon Web Services can offer – Aurora.
Amazon’s Aurora MySQL is advertised as a fully managed, drop-in replacement for MySQL with superior performance and high availability features. In this presentation you will hear from DraftKings’ de facto DBA. Learn from our experiences (and growing pains) with Aurora and whether or not the product lives up to the hype.
Dan White – Draftkings
GETTING TO THE CONFERENCE
Federal Reserve Plaza is the iconic silver “washboard” building located in the block defined by Summer Street, Atlantic Avenue, Congress Street, and Dorchester Avenue, near South Station and Dewey Square, at the start of the Rose Fitzgerald Kennedy Greenway.
The Boston Fed is easily accessible by the MBTA’s Red and Silver Lines. Take the Red or Silver Line to the South Station stop. The Bank is the tall silver building located at the corner of Summer Street and Atlantic Avenue. Note: The Silver Line provides service to and from Logan International Airport. For more directions including driving and parking, click here.