Hey, I'm Siddhant!
I'm an undergraduate currently at Princeton. In my free time, you can find me drawing maps or dancing, but when I'm studying, I'm probably learning more about artificial intelligence, software, and our planet's interconnected systems. I have a sister 11 years younger than me, and I love traveling... I can't sit still for long periods of time.
are data centers really that harmful?
"don't data centers use less water than golf courses? why are politicians vocal?"
🤖
key considerations
* imposed cost burdens on consumers
* energy grid constraints
* biome-relative resource depletion
* direct to chip vs ambient cooling
* gpu allocation inefficiencies
* on-site generation & growth of natural
gas resources vs decarbonization
* size, scale, networking & routing
* policy & a willing government
🏜️
coming soon (thesis)
do carbon markets reduce emissions?
the carbon market is a shared responsibility mechanism,
-
putting dollar value on carbon emissions to deter pollution.
-
placing a market cap on carbon emissions & encouraging firms to buy and sell credits (the right to emit)
the issue is most carbon markets are voluntary and unregulated, making it easy for firms to exploit market conditions & game the system.
deep machine learning (LSTM models) allow for anomaly transaction detection & reveal the economic variables and mechanisms that require market regulation.
read the full paper here! (junior paper)
how do renewable energy companies make money?
key contributions to clearway energy's energy forecasting method:
-
developer & product owner: new forecasting pipeline
-
researcher: do transformer-based deep learning forecasts outperform physics-models?
-
consultant: do clients' weather forecasts & models increase accuracy? leading client meetings
key outcome:
-
identified combined $1M+ monthly in recoverable revenue across two pilot sites.
a machine learning future.
hydromodel - predicting wildfire risk through a novel groundwater approach
-
groundwater as a non-traditional input for predicting wildfires
-
American Geophysical Union keynote (doi) and Princeton High Meadows
model selection ML - predicting waterfall forms to inform flood & erosion risks
-
erosion as a lens for hazards & disaster preparedness
-
NASA & Stockholm Junior Water awarded student research
digital mapmaking for a sustainable tomorrow.
Princeton GIS -
-
60 Seconds: Tsunami Evacuation Towers for a Cascadia Event
-
FRS 141: Encroachment on Indigenous Yanomami Lands
-
URB 385: Resilience Gentrification
WA Ecology Dept ArcGIS -
-
Washington's Hydrological Gem: Fjord Estuaries
-
Giants of the Puget Sound: the Southern Resident Orca Whales
🌲
& more
Princeton Admissions - a week in the life
More Projects - a vitae
Princeton Integrated Groundwater Modeling Center - Water and Climate Academy
Princeton Science Olympiad