ai · bioinformatics · research · building things
siddhartha
i build intelligent tools at the intersection of code, biology, and curiosity.
about
i like working where biology, computation, and design overlap. lately, that has meant biomedical AI research, gene-network thinking, and small tools that make complex work easier to understand. i'm especially interested in software that helps people move from messy information to clearer questions, better decisions, and useful action.
currently
experience / education
experience
research assistant
biomedical ai research
2025 - present
building machine learning pipelines for biological systems, including EGFR mutant drug-binding prediction and colorectal cancer gene-network analysis.
computer science tutor
University of Missouri–St. Louis
2024 - 2026
supported students in programming, algorithms, databases, machine learning, and systems courses.
student leader
campus + community organizations
2024 - 2026
led student communities, cultural events, technical workshops, and cross-campus collaborations.
education
m.s. artificial intelligence
Columbia University
new york, ny
aug 2026 - present
focusing on biomedical AI, machine learning, probabilistic modeling, and intelligent systems.
b.s. computer science
University of Missouri–St. Louis
st. louis, mo
jan 2023 - may 2026
studied computer science with work across AI, ML, software engineering, databases, systems, and research.
research
egfr mutant drug-binding prediction
building pipelines to study EGFR mutants and ligand-target binding behavior using biochemical affinity data and model benchmarking.
colorectal cancer gene networks
using heterogeneity-aware gene network analysis to uncover disease-specific expression patterns across colorectal cancer datasets.
biomedical ai
developing models and tools that make biological and clinical data easier to analyze, evaluate, and use.
agent context + memory engineering
building inspectable agent loops that separate context, state, and memory, gate tools behind policy, and evaluate the full system rather than only the final answer.
probabilistic machine learning
exploring gaussian processes, uncertainty, and optimization for scientific modeling.
projects
biomedical ai
TumorScope
an AI copilot for exploring tumor ecosystems from public spatial transcriptomics data.
notes
agents do not need to remember everything
what building a context + memory lab taught me about state, forgetting, tool safety, and the difference between useful memory and accumulated noise.
read notewhy i'm building at the edge of ai and biology
may 2026 · 4 min read · why biological systems, machine learning, and software design keep pulling me toward the same kinds of problems.
what makes a useful scientific interface
may 2026 · 4 min read · notes on designing calm tools for complex scientific workflows.
resume
resume
a concise overview of my work, research, and projects.