Professional Experience

Data Scientist

FewsLab, The George Washington University (June 2024 – Present)

I work at the intersection of data engineering, machine learning, and public policy - building systems that turn messy, large-scale data into research that matters. As a Research Data Scientist in the School of Engineering and Applied Science, I support federally funded projects across sustainability, energy, food systems, and civic infrastructure, partnering with agencies like NSF, USDA, and foundations like the Patrick J. McGovern Foundation to translate complex datasets into actionable insights.

Key contributions:

  • Modeled the U.S. nitrogen and phosphorus supply chain across all 50 states using 7+ integrated datasets on crop production, livestock, and interstate trade - with a deep focus on Chesapeake Bay counties to support the USDA’s 2030 and 2050 nitrogen reduction targets. Work delivered as part of the $8.97M USDA-funded “Thriving Agricultural Systems in Urban Landscapes” initiative and deployed at nitrogen-project.fewslab.org.
  • Engineered a national climate data repository from 66,000+ NREL files spanning 11,000+ U.S. energy plant locations under the NSF-funded “Ethical Implications of Connected Critical Infrastructure” ($509,880) and “GET North” ($249,997) programs - now powering renewable energy forecasting and infrastructure planning research with a research paper (Kollar et al.) currently under review.
  • Built a UK Parliamentary NLP pipeline processing 670,000+ records from the ParlaMint Commons corpus (2015–2022) using BERT-based models under NSF support - published as Robertson et al. (Sept 2025) .
  • Evaluated the Healthy Corners urban corner-store program in Washington, D.C., in partnership with DC Central Kitchen and the Partnership for a Healthier America, under a $500K Patrick J. McGovern Foundation grant - building data infrastructure to measure healthy food access in underserved neighborhoods.
  • Conducted causal inference on Vision Zero NYC using 2M+ crash records, applying difference-in-differences with parallel-trends validation. Presented as a poster at AGU 2024.
  • Built interactive dashboards in Python Dash and R Shiny to surface patterns across 100K+ records, enabling faster decision-making for researchers and stakeholders.
  • Applied K-means, DBSCAN, and LSTM models to forecast NYC taxi demand and detect pricing anomalies in 20+ years of NYISO electricity market data.
  • Designed automated ETL pipelines and custom web scrapers that cut data preparation time by 80%, freeing weeks of manual effort.
  • Mapped executive networks from 10,000+ SEC filings using Neo4j, revealing influence patterns across 1,500+ corporate leaders.
  • Built a multimodal misinformation detection system fusing ResNet-18, CLIP, BLIP, and BERT to validate image–caption alignment on a 10K+ image dataset from a National Forensics Department - achieving 95% accuracy.
  • Published an open dataset (doi:10.5281/zenodo.17372375) with an accompanying Python package examining the effect of U.S. data centers on climate-related public concern.

District Data Coordinator

Inqui-Lab Foundation — sponsored by UNICEF and the Government of Telangana (Oct 2022 - July 2023)

I owned the data strategy for the School Innovation Challenge 2022 - a statewide initiative to spark creativity and problem-solving in public schools across India.

Key contributions:

  • Tracked and analyzed real-time student performance data across 7 districts, driving a 35% increase in participation and reaching over 64,500 students.
  • Built automated data pipelines in Python that eliminated 20+ hours/week of manual processing - letting the team focus on identifying where intervention was needed most.
  • Designed dashboard reports and custom visualizations to monitor engagement, performance, and resource allocation - directly supporting 4,900+ teachers and District Science Officers in making faster, data-informed decisions.
  • Developed feedback collection tools that shortened decision cycles for educational improvements at the grassroots level.

Education

M.S. in Data Analytics The George Washington University - Graduated December 2025

B.Tech in Computer Science Jawaharlal Nehru Technological University, Hyderabad

Selected Publications

  • Robertson et al. (2025). UK Parliamentary discourse analysis using BERT-based NLP on 670K+ ParlaMint Commons records (2015–2022). Published September 2025.
  • Kollar et al. . Under review.
  • Rao et al. (2023). AIP Conference Proceedings.
  • Kumanduri, T. (2025). Data Centers’ Effect on Climate-Related Concerns in the USA — open dataset and Python package. doi:10.5281/zenodo.17372375

Hackathons

  • Hoya Hacks 2026 — Scene-Speak: an LLM-orchestrated multimodal application.
  • George Hacks 2025 — My Own Medic: an AI-powered health chatbot.

Volunteering & Leadership

Volunteer, The Art of Living Foundation

(2019 – Present) As a certified yoga and mindfulness instructor, I’ve facilitated personal development sessions and wellness workshops for over 1,800+ participants — from university students to corporate teams. Sessions have been hosted at USC, GWU, Babson College, IIT Hyderabad, and NIFT Hyderabad, centered on mental well-being, focus, and resilience.

Finance Director, GW Desis

(Jan 2024 – Present) I manage financial planning and budgeting for a 350+ member graduate student community at GW - organizing cultural, professional, and networking events while keeping resources lean and impact high.

Digital Marketing Director, Google Developer Student Club (GWU)

(Mar 2024 – Present) Spearheaded outreach and marketing for GDSC workshops and tech events, drawing 150+ students per session. My focus: making Google technologies accessible and building a culture of peer-driven learning on campus.

Summary

I care about building things that matter - whether that’s a pipeline processing 66,000 climate files for the NSF, a nitrogen supply-chain model supporting USDA’s Chesapeake Bay targets, or a wellness workshop helping students manage stress before finals. These roles reflect a through-line of using data, engineering, and leadership to create impact that’s both meaningful and measurable.