Abhijith (@abhijith0505) is a Staff ML Engineer at Target, driving architecture and standardization of large-scale personalization systems and ML platforms. This is his personal website where he writes about ML systems, platform design, and scaling ML in production.
I cannot pinpoint exactly when this started. At some point over the last two years, coffee stopped being background noise and became something I was actively paying attention to. I went from “just give me something with caffeine” to having a fairly specific answer to the question — what do I actually want to drink, and when? This is my attempt to document that. What I Drink and When There’s a…
I was talking to a friend the other night — someone who spent five years in the tech industry before deciding to move into academia. We started out chatting about random things, but at some point, the conversation drifted toward work, purpose, and what it means to spend your life in front of a screen. He said something that stuck with me: “Dunno, feels kinda odd to be thinking of going back and…
Introduction If you’re considering the Online Master of Science in Computer Science (OMSCS) from Georgia Tech, you’re not alone. It’s one of the most well-known and affordable online master’s programs in CS, and since finishing my Machine Learning specialization, I’ve had a ton of people reach out with questions about admissions, coursework, workload, career impact, and whether it’s actually worth…
December 2021 - Four years into my first job, I asked my manager for the opportunity to mentor at least one intern. I wanted to be a good mentor. Unexpectedly, at the start of 2022, I moved teams and began working with four people. I was appointed as the tech lead for a group tasked with building recommendation systems focusing on ‘exploration’ (addressing cold-start and sparse data problems).…
Fin. I have completed my Master’s program from Georgia Tech’s Online Master of Science in Computer Science (OMSCS). Phew. It was a journey, alright. For many who complete this program, the final course is often Introduction to Graduate Algorithms, commonly known as GA. It’s a mammoth of a course, and reviews from OMSCentral consistently classify it as the toughest course for the ML specialization.…
When processing big data, efficiency is key. It’s not uncommon to be caught up in long debugging cycles when working with Spark. I was recently caught in such a debugging train when one of my pipelines was taking longer than expected. It was a simple structured streaming pipeline that was listening to a Kafka topic for events and performing some transformations on it, before emitting the output to…
In the Fall of 2023, my journey through OMSCS led me to explore Game AI (CS7632), along with HCI, a course that promised excitement and a bit of a departure from my usual focus. Having a blast in AI for Robotics (AI4R), I thought, “Why not try something new?” Enter C# (never touched it), Unity Engine (never installed it), Game development (complete newbie), and, of course, AI (my comfort zone).…
As I inch closer to the culmination of my academic journey, I selected my courses for this semester aiming to strike a balance between relevance and manageability - Human Computer Interaction (HCI) and Game AI (GAI). The former was intended to be a light but pertinent addition to complement the ML specilisation that I am pursuing. So, HCI wasn’t about coding, but more about papers. The lectures…
In the first installment of my reading saga, I spilled the beans on how I resurrected my reading habit, danced through genres, and discovered the art of balancing fiction and non-fiction. I’m back with the sequel – think of it as the Empire Strikes Back, but with fewer lightsabers and more books, with some focus on what went right and what was troubling. Revisiting the Grand Plan - The core of my…
As the summer sun lit the horizon, I enrolled in the Software Development Processes (SDP) course. This choice was driven by the tantalizing prospect of refreshing my Android Development skills, a requirement for my new role at work where I am tasked to put recommendation systems on the edge mobile devices - federated learning stuff (more on this later). In terms of academic load, SDP proved to be…