Software engineer Sam Kinard is no stranger to Austin tech. Having worked for numerous Austin firms and startups, including Spredfast and Bazaarvoice, he’s spent much of the last decade building systems that increase customer engagement.
While these opportunities provided him with fulfilling work and set the groundwork for lifelong friendships, Kinard said when he discovered what CCC’s local office was up to two years ago, he couldn’t miss out.
Having joined CCC initially as a senior platform engineer, today Kinard leads CCC’s Austin-based engineering team as manager of software development. Together, they’re building blockchain, data science, hyperscale computing and machine learning solutions for the auto industry with the end goal of integrating AI into every single one of CCC’s products. It’s a lofty goal, but Kinard and his team are up for the challenge and already making waves.
What inspired your move to CCC a couple years ago?
CCC is pushing the envelope across the entire automotive industry. We are building bleeding-edge products and platforms in artificial intelligence and machine learning. We are trying to accomplish a lot of things in these areas that no one has been able to do yet. It’s very exciting and scratches the itch for something new and different.
Talk to us about the problems you solve when you come into work, and the technologies you're using or building to accomplish those goals.
One of the main objectives of our team is to bring AI to all CCC products — both existing and future offerings. A big technical challenge we face is taking new AI models created by our model development team and running them in a robust, large-scale production environment. Running a model on a laptop or in a development environment is very different from testing it within a system that must handle hundreds of thousands of transactions per second. We’ve developed a number of platforms over the past few years to accomplish this, using technologies such as Spark, Storm, Itsio, Tensorflow, Kafka, and Kubernetes. Additionally, our group has used “works out of the box” frameworks such as AWS Sagemaker when appropriate.
Our group is building an AI pipeline that goes from raw data to production application and allows everyone involved to do their best work without interfering with each other.”
What’s the biggest technical challenge you’ve faced at CCC and how did you overcome it?
Our data science and research and development teams work at a breakneck pace, constantly evaluating and utilizing the newest open source technologies and the latest machine learning algorithms. These products quickly transition from a proof of concept to a production environment with strict requirements for reliability and performance. In many cases, the ability to rapidly produce new AI products using the latest experimental technologies is directly at odds with providing a stable, performant environment for production customers. The answer to this is obfuscating disruption through abstraction. Our group is building an AI pipeline that goes from raw data to production application and allows everyone involved to do their best work without interfering with each other.
What technical challenges lie ahead and how will new hires help address those challenges?
The effort to build out the AI pipeline will continue to be a huge focus of our team and the larger organization. Many slices of what the pipeline does can be accomplished using tools and frameworks that have come out in recent years. However, no one has yet been able to put everything together in such a cohesive way that bleeding edge models can be rapidly developed and then deployed to a robust, scalable production environment. What we are trying to do is very ambitious and difficult. We need a lot more folks to come on board and help get it done.
Outside of technical skills, how else do you identify if someone will be a good fit for your team?
We have a close-knit team of folks who love working together, which is something we are very proud of. We look to our team members to be vocal during group discussions while also hearing and considering the dissenting opinions of their peers. We highly value the ability to pick up and learn new technologies, skill sets, or disciplines that are unfamiliar. We also look for people on our team to be proactive with their work. If they feel stuck, or have completed their current work, they should touch base with others to figure out what needs to be done, rather than waiting for someone to come check on them. Someone who can anticipate the future needs of ongoing projects would be a great addition to our team.
What is something you think interested candidates should know about your team that they might find surprising?
We are an AI team that didn’t start with a background in AI. A few years ago we didn’t have much experience with AI across members. We set aside some time for everyone to get up to speed on the subject. Folks were able to utilize whatever learning style they preferred. We even organized some small hackathons and contests to make things more interesting. Today, we have one of the few teams known to be AI experts within CCC.