Now. Present. Alive. I guess…
Mostly staring into the abyss. Machine Learning + Human Learning + Existential Crises (pondering about knowledge, learning, life, universe and everything.)
(Last updated September 7, 2024)
You can call me whatever you like though. I am a paradox!
Currently, I am working as a “Computer Scientist” at NASA-IMPACT.
I am at the intersection of Machine Learning research and engineering with NASA-funded projects, particularly focusing on Earth Science. My role involves delving into research problems, generating new ideas, and applying existing ones to solve challenges that advance our understanding of Earth systems.
I’m a strong advocate for open science, and this aligns well with NASA-IMPACT’s mission to push boundaries in Earth Science through machine learning.
Some works include building tools like:
I am currently working in Natural Language Processing/Understanding(NLP/NLU), especially with Large Language Models (LLMs), leveraging their capabilities to empower machine learning workflows for various NASA projects. My current work involves – research + engineering – LLM training, fine-tuning, active learning, and implementing guardrails to ensure reliable model outputs. Additionally, the work also involve low-resource ML solutions where we’re dealing with low amount of data and smaller language models.
As a Graduate Research Assistant (GRA), I focused on applied research for my master’s thesis, Feature Selection in High-Dimensional Space with Applications to Gene Expression Data. I collaborated with NASA’s Bio-Physical Science (BPS) team to develop machine learning-based gene ranking algorithms for space-flown rodents’ gene expression data.
I presented my poster at AGU 22, which highlighted my ensemble-based ranking framework to identify significant genes contributing to physical attributes such as gender, age, and condition of the rodents. Final Slide Deck.
Beyond that, everything is in the past now, shaping my current work and philosophy as a researcher and a human being.
Additionally, I also love to actively engage in workshops and presentations to share insights on machine learning and LLMs, and to foster innovation in the Earth Science and AI community.
NASA SMD AI Workshop (March 25-27, 2024): Volunteered and gave a talk/presentation on Evaluating LLMs in the context of Earth Science applications. This workshop brought together experts from various domains to explore the role of AI and ML in advancing NASA’s Science Mission Directorate (SMD) goals. Workshop details and speaker page.
SRM University Hackathon & Workshop (April 2024, India): Conducted a workshop and organized a hackathon focused on leveraging machine learning workflows for remote sensing applications. This event aimed to encourage students to apply ML techniques to real-world challenges in Earth Science. LinkedIn post about the event.
These experiences reflect my passion for knowledge-sharing and fostering a collaborative environment for innovation in AI and Open Science.
I am Nish. I don’t do much, except stare into the abyss and try to make sense of this paradox we call existence.
BTW, I mostly document my thoughts in my personal knowledge base using logseq, which is open-source! It’s a bit chaotic, but it helps me stay grounded.
Vipassana Retreat: Oh I also went for a 10-day Vipassana retreat at Idaho Falls in mid-July 2024. Whenver I am asked the question “how was it?”, I reply, “It was transformative”. :) I try to meditate whenever I am free of other works.
I once started a newsletter that feels like a piece of the ancient past. Here it is: bits-and-paradoxes.
Life is a suffering-optimization algorithm!