I’m a Northeastern M.S. in AI graduate specializing in computer vision, verifiable ML, and robustly controlled multi-agent AI systems. I built end-to-end CV at Juhomi, developed AI microservices rapidly at Sentient, addressed facets of AI in healthcare at Inflohealth, and engineered zero-shot computer vision workflows and multimodal agentic systems with rigorous testing at Norfolk Southern.

the candid version:

• I started liking AI because it seemed magical. “You can make machines see the world like you do!, what’s not to like?” - younger me.

• As I started learning more, turns out, it’s made possible using elegant math and code! BAM! [statquest]

• I was lucky to work on complete end-to-end computer vision pipeline at Juhomi (2021-2022). That was like being overwhelmed at the sight of the ocean, just as I was amazed after discovering water. Whiplash! never stopped tho.

• I began to get a groove during my work with Sentient(2023), met a couple AI ninjas, and learned some ninjutsu myself.

• Then I came to the USA, to Northeastern for my Masters in AI. before I was overwhelmed by a peek, and now, this was like cliff-jumping into the ocean 🏊🏽‍♂️.

• I was now clearer than ever about I wanted to do. My view of AI when from “imbuing computers with magic 👶🏽🪄” to “Recreating physics with data 👨🏾‍🔬👨🏾‍💻”

• As time here progressed, My mind was being opened to the different facets and challenges in the field. My time at inflohealth(2024) dealth with the many practical concerns of AI, and it’s application in critical domains (like healthcare). It’s not always magical😅, but it still is🙃.

• As I took classes like Verifiable ML, it became clearer that AI, if wielded irresponsibly, will trash your system at best, and a whole lot of others’ at worst. My distaste for anything Prompt-and-pray-flavoured was higher than ever.

• Working with the Data Science team at Norfolk Southern, I was reconnected with one of my core passions COMPUTER VISION!. Working with their crazzy tech! I didnt notice time going by as I GEEKED OUT!. As I began working on multimodal multi-agent systems, building testing workflows that were more robust than ever to keep the agents in check was the biggest challenge I had ever faced. I had never appreciated how critical thoughtful software engineering is more than I did when I was here.

• As I finish my master’s, I feel a twinge of regret for not seeing earlier how profound math and science could be; seeing them as I do now might have left me better off. It’s never too late tho 😉🤫😶‍🌫️.

• Lately, I’ve been exploring what it means to design Human-Centered AI systems. If there is one thing AI systems today need, it is adherence to Human-centered design principles. Proficiency alone doesn’t foster adoption; Reliability, Safety and Trustworthiness do.

• Can’t wait for what’s to come!

👉 You can explore my portfolio, GitHub, and LinkedIn.