October 31 - November 2  - Portland, OR
Click Here For Information & Registration
View analytic

Log in to bookmark your favorites and sync them to your phone or calendar.

API Success Stories [clear filter]
Thursday, November 2


How We Built a Thriving API Ecosystem at a 65-year-old Company - Rick MacConnell, Cox Automotive
Manheim is a 65 year old company and is the largest automotive wholesaler in the world. The company has used the IBM iSeries platform since the 1970's as a core part of it's brick and mortar business. Two years ago, we set out to transform the technology foundation at Manheim with a goal of accelerating developer productivity and flexibility. We started by establishing a vision for a developer-centric API alongside a cloud vision. The first system that we built uses Apache Camel which listens to streams of changes coming from each of the individual iSeries boxes at our brick and mortar locations and turns these changes into logical business events. Camel then routes these events to a series of Java Spring processors that handle them and deliver them on to interested subscribers. This system has been a critical addition and has transformed the way we build software. It has led both to our successful ability to integrate with third parties as well as being a core part of our ongoing ERP implementation. Following this, we developed an API cookbook that prescribes rules for the structure and behavior of our REST based entity centric API. Using this cookbook, we developed a foundational API catalog that gave developers outside of Manheim the ability to interact with key services including inventory and wholesale pricing. At the same time, we committed to creating a robust developer portal with a commitment that our API should pass the 15 minute test with developers. Our goal is that they should be able to go to our portal, read through the documentation and get code up and running in 15 minutes. Finally, building on the success of both our real-time event delivery system and our API, we built a system called Eventer. Eventer is a cloud native system that lets our partners use an API to establish subscriptions to events that they're interested in receiving. Using a variety of cloud SAAS tools and Java Spring code, we've built a self-service API driven system that allows third parties to subscribe to and receive events in real-time. Combining the three concepts; an API vision, event-driven data and the developer portal, we've built a thriving ecosystem that has taken on a life of it's own. We are now having to develop new support structures in order to service the growing customer base that is adopting our API. Additionally, we've succeeded in transforming the way we build software here. One of our internal partners used our real-time events to roll out a new business model across the country in three months. This would not have even been conceivable prior to the existence of our API.


Rick MacConnell

AVP Engineering, Cox Automotive
Rick MacConnell is a software developer who has worked on large-scale enterprise applications for the last 20 years. He has been with Cox Automotive for 16 years where he started working on AutoTrader.com. Over the past 3 years, he has been focused on building an API ecosystem at... Read More →

Thursday November 2, 2017 2:00pm - 2:20pm
Atrium BR


REST API Microversions - Sean Dague, IBM
OpenStack ran into a challenge: how do you balance the ability to evolve a REST API, with backwards compatibility guarantees, and have it work the same across different deployments, when you have no control over when they are going to upgrade. The answer we came up with was labeled "Microversions", inspired by HTTP content negotiation. This talk will dig into the challenges of versioning REST APIs, as well as the unique challenges of versioning ones that expect to exist in interoperable public clouds. It will dive through ideas that were tried and retired in OpenStack, and describe in depth the microversion system that many OpenStack services now use. We'll look at what's been learned in 2 years of this in the field, and suggest how related approaches might be applied to other open source projects.

avatar for Sean Dague

Sean Dague

Software Engineer, IBM
Sean Dague has been an Open Source developer for most of his professional life. He's worked on numerous Open Source projects over the years including SystemImager, OpenHPI, Xen, OpenSim, NFS Ganesha, and OpenStack. | | He's a core reviewer on Nova, Tempest, Devstack, Grenade, and... Read More →

Thursday November 2, 2017 2:30pm - 2:50pm
Atrium BR


How We Doubled the Velocity of Our Developer Experience Team - Matt Bernier, SendGrid
We ruthlessly prioritize everything we do and automate away anything we have to do more than a couple times. This has helped the SendGrid DX team to maintain libraries for 7 different programming languages and their github communities with a single engineer. We challenge ourselves to always be working on the highest impact items for our users. Our goals are to maintain engagement with our developer community, manage our libraries at the highest possible level, all while sustaining our lean team.

avatar for Matt Bernier

Matt Bernier

Developer Experience Product Manager, SendGrid
Matt Bernier is an avid proponent for maximizing the experience of developers with APIs. He gardens, bakes things, cans things, brews things, fathers things, manages products, and uses sriracha as if it were ketchup. He reads 50 books or more a year. On a business trip to Beijing... Read More →

Thursday November 2, 2017 3:00pm - 3:20pm
Atrium BR


Contract-first API Development: A Case Study in Parallel API Publishing & Consumption - Chris Busse, APIvista
The idea of designing an API first, then having both the publishing developer teams and consuming client developer teams do their work in parallel is a pattern that the standards & tooling of the Open API ecosystem make worth considering. However, this approach is not without its challenges especially when it comes to governance and communication.

This session will present a real-world case study where this approach was executed successfully, going over how the work, teams, and deliverables were structured and presenting lessons learned that will be useful to anyone else considering this working model.

avatar for Chris Busse

Chris Busse

CTO, APIvista
Chris Busse works at APIvista as CTO where he helps his clients design, publish, and support APIs. Before joining APIvista he lead the API Consumer Services team for the Capital One DevExchange platform, supporting internal and external business partners while developing new product... Read More →

Thursday November 2, 2017 3:30pm - 3:50pm
Atrium BR