My Impact as a Summer Intern at Optimizely

I heard this phrase numerous times throughout my onboarding process at Optimizely. But coming in as a software engineering intern this summer, with every tech internship program out there being “hands-on” this, “impactful” that, I wondered if I would be remotely close to any customers, let alone making them wildly successful. It was my first real job. And as someone with relatively little experience, at best, I expected to be handed an interesting starter intern project; at worst, relegated to the dark fate of doing… “intern” work.

I learned quickly that Optimizely’s commitment to “making our customers wildly successful” does not discriminate against position or experience. My first main project on the Data Warehouse team this summer was one of the major initiatives for the Customer Success team: to build the Mavenlink data pipeline for the Customer Success team.

Optimizely enables our customers to turn data into action. In a similar vein, the Data Warehouse fetches and stores data from the various product and business systems that Optimizely uses, so that we can leverage our own data. Data is no longer locked up just inside third-party systems, but is accessible from our own data store. With over thirty data sets in it, Optimizely employees are truly empowered to derive non-obvious business insights from these rich data sets.

Mavenlink is a system our Services team uses to record service engagement hours with our customers on their projects. This would include information such as billable hours, project completion rates, and customer support engineers involved in the work etc.

While Mavenlink’s out of the box analytics provide a lot of information on how our Services team is doing, it does not give a comprehensive view of other dimensions associated with customer’s product usage and other attributes of the customer. This inhibits our ability to do comprehensive analytics on our Services engagement and it could potentially limit our ability to better understand and serve our customer needs.

Generally, a data pipeline involves an ELT (extract, load, and transform) process. Extract the data from the source, load it into a temporary staging area and then, transform the data into a neat OLAP schema format so it’s easily accessible for Optimizely employees. Building Optimizely’s Mavenlink data pipeline involved 5 main steps:

The most time-consuming part of the project wasn’t the actual development work, but reading through the codebase to figure out what’s in a pipeline, and understanding the overall architecture. The problem was that all these data sources had different APIs and schemas, so finding reusable code, patching that together and writing what I needed, combined with getting used to Docker (a container software) and Phabricator (a code review tool for github), and learning the ropes of development procedures, took some time.

It really was an eye-opening experience. I learned a lot about data pipelines, but more importantly, about coding practice and soft workplace skills. As the initiative was closely tied with the Services team, I learned the value of double- and even triple-checking.

Finally, I put the Mavenlink pipeline in production to run every hour. I’m happy to say that it’s been running smoothly since its rollout!

Having Mavenlink data in the Data Warehouse gives immense ability and power to our Customer Success team to do analytics way beyond what Mavenlink’s out-of-box functionality can provide. This enhanced power mainly comes from combining services data with the product metrics and customer attributes that are not available in Mavenlink. With increased focus from Optimizely on the Services wing, it is very important for Optimizely to accurately measure the effectiveness of our engagement with our customers and prove the value.

Besides the meaningful projects that I got to do, the people at Optimizely added a whole other dimension. Everyone in the company was always open to questions, conversations, or coffee, and super generous with feedback and recognition. In particular, I never felt any hesitation at all to ask my mentor or teammates for guidance.

So… I guess you could say that I really optimized this summer!

Related posts:

In Amal academy we are a lot of new things on the daily baises. The challenges are becoming more and more interesting. Last week I learned 6 tips mentioned by our mentor Sohail Sb. Regarding…

Here are two things that are almost always true about heterosexual men — they watch porn periodically, and they fantasize about having sex with a porn star at some point in their life. I’m not really…

In the last few years we’ve made huge strides in recognising, diagnosing and talking about Mental Illness. That has to be a good thing. It helps us find the help we need and it helps us heal. The…