Our own Avantika (Avan) Mathur will be giving a talk at this month's Merge 2016 conference in San Francisco on Strategies for Managing Microservices at Scale .
Our enterprise customers all know Avantika very well. Now, her upcoming talk is the perfect opportunity for everyone to benefit from her vast technical experience tackling some of the toughest problems with today's large-scale, complex, delivery pipelines. Ahead of her talk, we sat down with Avan for a quick Q&A to introduce her to our readers.
QWhat is your role at CloudBees?
As Global Technical Account Manager at CloudBees, I work with some of our largest customers, and help them use our solutions to improve their software development and delivery processes. I help with overall strategy deploying our software across large organizations, design and implementation, as well as looking ahead to make sure the solution support future growth plans.
Q What is your background?
I grew up in Portland, Oregon and attended Oregon State University where I received a BS in Computer Science. Before joining CloudBees, I worked as a kernel developer at IBM for five years. I now live in Orange County, California with my husband and dog Scooter.
Q What do you like most about your job?
I love learning about my customers’ goals and strategies in DevOps, and developing solutions using our technology to help them succeed. Each customer faces different challenges, with new tools and processes to learn and work with to create a customized solution for their specific needs.
Q You help large customers solve huge problems – can you share a bit what those problems/challenges are?
My main focus - and passion - is working with those customers that really push the envelop in terms of the scale they need to support across the entire organization. I have customers that have 40,000 developers using our solution for build/test, others that are doing more than 10,000 releases per year , 100,000 builds per day , 1 million integration tests per year , or hundreds of millions of test cases per day (!).
When working with customers that need to support such a large scale, it is important to look ahead and plan for the expected volumes in the future. In the beginning, everything is small and works well, but we need to be ready for when you grow quickly and stay ahead of impact of the expansion. Aside from designing implementation to scale, a big part of this is making sure you have theright infrastructure to support automation at scale as well as the right pipeline design and workflows to ease on-boarding and support scaling across multiple teams or applications. Designing a pipeline for releasing one application might be simple enough, but what if you need to on-board 20,000 applications within a quarter? and do it in a standardized, optimized way, across hundreds of teams?
I work with new customers to help them plan for scalability, or help existing customers manage the scale that they were not ready for. It's exciting to see what their new performance metrics would end up being, and then planning to crush those time and time again :)
Such an intense requirement for scalability on all levels - the design of the pipelines and tool chain, processes, run time, infrastructure, etc. may seem daunting, but from my perspective it is extremely motivating. I see Scalability as a "good problem" that our customers have - where they're doing so well that they need to continue to accelerate and scale their software production. I take pride in the fact that our solution is proven at such intense enterprise work loads, and that it is so critical for their business and day-to-day operations.
In addition to designing for scalability, I also often help customers implement pipelines that integrate new technologies - such as Microservices and containers, or support modern deployment patterns, such as blue/green or feature-flag deployments.
QWhat is one of the achievements you’re most proud of that you’ve helped your customers:
I developed a self service cloud portal to automate spin up, configuration and delivery of cloud resources on demand, built on our CloudBees Flow platform. Using the portal, wait time for engineers to get a cloud resource went from weeks to just minutes. This was a fun and challenging project where I worked closely with the customer and saw great results from our collaboration.
Q What are you working on today?
I am working with one of our largest customers architecting a scalable process for deploying applications across their organization, and designing the architecture to support it. We are modeling the solution for scale and accelerated adoption. It is a big undertaking, but we have a great team working together.
QWhat do you like to do in your spare time?
When I’m not traveling to customer sites, I enjoy hiking with my husband and friends, spending time in my garden, or curling up with a good book and a cup of tea.
If you're attending Merge 2016, April 13-15 in San Francisco, be sure to catch Avan's talk on Strategies for Managing Microservices at scale, Thursday 3:40-4:10 p.m. - Ballroom III - Level 4.
Avantika will cover tips for evaluating if Microservices are right for your particular set-up; how to get started if you wish to decompose your Monolithic application into microservices; and some best practices for designing your software delivery Pipeline and Operations framework to operationalize microservices at enterprise-scale. Using a real case study and a live demo of Perforce Helix, Docker and Selenium, Avantika will be sharing patterns for enabling a scalable and extensible pipeline to support today’s modern app delivery with microservices.