This as-told-to essay is based on a conversation with Adwait Sathe, a 32-year-old senior data privacy engineer at Meta Superintelligence Labs. It has been edited for length and clarity.
I’ve been working at Meta for the last five years, first at Messenger, and for the last three years as a data privacy engineer at Meta Superintelligence Labs.
I moved to the US from India in 2019 to pursue a master’s degree in information systems at Northeastern University in Boston, graduating in 2021.
I knew the job application process out of college would be competitive, so I started my job search about a year and a half before I graduated.
I was very clear from the start: I want to work in data engineering privacy. I took classes in that field and built a relevant portfolio through internships and side projects.
In my second semester, I started looking at job postings I would apply for a year later to understand what the industry was looking for.
I had four offers in hand when I graduated — from McKinsey, Walmart, Amazon, and Meta. Here’s the strategy I used to approach my job search.
Cold messaging people who weren’t too senior
I had a tactic for cold messaging people on LinkedIn.
I connected with a lot of people on LinkedIn who were working on things similar to what I wanted to work on.
The people I reached out to were active on LinkedIn and working in Big Tech in data engineering roles, not managerial roles. Engineers and senior engineers, not vice presidents or directors.
I was looking for mentorship. I approached them to understand the industry’s requirements, so that when I applied, my application would stand out among the masses.
My cold message was very simple: “Hey, I am Adwait, I’m studying this, and I want to pursue my career in this. I have this tech stack and résumé; what should I improve?”
One connection gave me excellent advice — that my résumé made it look like I wasn’t experienced.
I had three years of work experience in India after getting my bachelor’s degree. Still, when I first put together my résumé, it looked like a new grad résumé because I front-loaded my education over my work experience.
I started getting more interviews when I fixed that issue.
When I started cold messaging with this strategy, I aimed for volume, sending messages to a lot of people. Many didn’t reply.
But once I had gotten three or four good mentors, I eased up on messaging, knowing that these people were really solid and would help me out.
I’ve tried to pass it on. Now, I reply to almost all LinkedIn messages I receive that ask for career advice.
Splitting job apps by volume and strategy
My second job search strategy was to split my application process into two buckets.
The first bucket: companies I really wanted to work for.
I had a list of companies I wanted to work for, and I used to go to their career websites a couple of times a week to check for new positions and apply early. I would be looking out for new grad positions, which are typically released way before graduation season.
The second bucket was wider. This was all about volume, applying to as many roles as possible.
I had a generic résumé for all the volume-based applications. But for the ones I really wanted, I tailored each application.
After I got my first offer, a data engineering role at Walmart, I stopped applying in volume. The strategy after that was to apply for roles I really wanted, which could match or exceed what I’d already gotten.
How I would apply differently today with AI
If I were applying for the same roles today, I would make sure to have a lot of different projects in my portfolio, because building has become so important.
A working product, which is live and has real users, is way better than bringing a prototype into the interview and saying, “I built this prototype, it worked when I built it, but it’s not working now.”
The second thing is AI fluency. This doesn’t just mean understanding different tools, but also how AI works in the background and a judgment of when to use which tools.
But the fundamentals don’t change. Being good at coding and knowing how to structure the code is still important. AI can handle implementation, but designing systems and architecture is still very important and human.
