Machine Learning and Lifelong Learning with Srivatsan Ramanujam

Srivatsan is a Director of Software Engineering, Machine Learning at Salesforce and was recently interviewed on the Breaking 404 podcast hosted by Hacker Earth. This is his story.

If he weren’t an engineering leader, Srivatsan Ramanujam would be a National Park Ranger in Yellowstone or Denali, much to the chagrin of his wife. Thankfully for us, he instead works in our San Francisco office as a Director of Software Engineering in Machine Learning helping to build customer-facing machine learning products within Sales Cloud Einstein and Pardot Einstein.

Read Parts 1 and 2 of the journey to machine learning driven sales and marketing!

Srivatsan received a Bachelors in Engineering from PSG College of Technology in India and worked as a software engineer for several years before heading to University of Texas at Austin (hook ’em Horns!) to pursue a graduate degree in Computer Science, specializing in machine learning, statistical data mining, and natural language processing. A role at Salesforce, based in San Francisco, was his first job after completing his Master’s degree. He then moved to roles at Sony and Pivotal before boomeranging back to Salesforce. Taking a trip down memory lane, Srivatsan says the first language he coded in was BASIC. He then transitioned to C, C++, and Java before going back to “ancient dinosaurs” like COBOL (hey, we still need it!) and IBM 360 Assembly Language before finally discovering what he deems “programmer’s Nirvana” — Python!

Of his background in data science, Srivatsan says he didn’t have formal training in data science or machine learning (ML) until his second job out of undergrad in India. That was when he started working for a company on computational drug discovery, building ML algorithms for a product that helped computational chemists build models to predict properties of drug molecules. This project inspired him to get more formal training in the field of data science, leading to his pursuit of graduate studies in computer science at UT Austin.

When asked about distinctions between the roles of machine learning engineer and data scientists, Srivatsan shared the following thoughts on the Breaking 404 podcast:

A machine learning engineer, he said, owns the end-to-end pipelines serving a product. This requires both software engineering skills and background in ML, including linear algebra and statistics, and also a general curiosity for data. From looking back at the innovations in software engineering over the past few decades that have made software engineers specialize, Srivatsan predicts that advances in tooling will make ML engineers focus on more advanced problems, as more and more routine tasks are automated.

As for the role of a Data Scientist, the nature of this job may be very different across companies. A data scientist may perform ad-hoc analysis of business problems and apply ML research to solve them. They may or may not operationalize their models. Data Scientists are usually not responsible for maintaining production pipelines, so they have the opportunity to be scrappy and iterate quickly to uncover interesting business insights.

Listen to the full podcast interview to find out what Srivatsan has to say about we handle the security and privacy of the data that is being used to train models at Salesforce and how we balance technical stability with delivering high-quality code!

Srivatsan is truly a lifelong learner who says he tries to attend as many of our internal talks as he can to hear from Data Scientists and engineers about their innovations. “I follow a lot of great people a lot smarter than I am on Twitter,” he says. “It helps me stay current on what’s the latest, greatest and most controversial in ML/DS.” He also uses exercise time (he’s a runner) to listen to podcasts; EconTalk and Data Skeptic are two of his favorites. He reads articles on Medium (thanks!) and browses through code repos on GitHub to stay up-to-date. When he’s able, he enjoys attending conferences focused on machine learning and data science.

His top tip for engineering managers and other leaders? Be genuine, and you’ll earn the trust of your team. He continues, “I believe in candor and being accessible to my team. Strong relationships are not only built in work-related conversations but also in how you connect on a personal level with your team. I try to never miss any lunchtime banter with my team.” As to how this has been affected by working remotely of late, he emphasizes the importance of over-communicating and using any and all communication channels available to you to stay in touch.


Image for post


We’re glad Srivatsan took the time away from his preferred leisure activities of hiking and backpacking to let us get to know him better on the Breaking 404 podcast! Follow him on Twitter to see what interesting things he’s learning and maybe catch some photos of great summit views. You can also subscribe to his blogs on


Image for post