03Σ experience

Research, systems, and computation.

Five roles spanning large-scale ML research, academic computing, product analytics, and GPU-accelerated scientific computing.

01 / Aug 2025 – Present

P(stance | narrative)

Graduate Research Capstone

LLM Stance Perception for Women’s Safety Narratives

Rochester Institute of Technology · Rochester, NY

Large-scale discourse and stance analysis across women’s-safety cases in India.

Built an end-to-end pipeline that reduced 4.3M+ raw Reddit and YouTube comments to a curated dataset of 351,501 comments across 16 cases from 2012–2024.
Used Mann–Whitney U, chi-square, and G-tests to compare stance distributions, including politically implicated and non-political cases.
Adapted Qwen3.5-9B with LoRA and evaluated it on a 3,000-sample held-out set: 0.7447 accuracy, 0.7450 macro-F1, and 0.6634 MCC.
Ran labeling, adaptation, validation, and systematic error analysis through RIT research-compute and SLURM workflows.
Co-authored the resulting short paper, accepted at ASONAM 2026.

PythonQwen3.5-9BLoRASLURMHypothesis testingNLP

02 / Aug 2025 – Present

uptime → 1

Rochester Institute of Technology · Rochester, NY

Academic computing support, software implementation, and resource analysis.

Supported faculty, staff, and students by maintaining academic IT and computing systems.
Implemented academic software applications with Java and Docker.
Monitored laboratory resources, analyzed system performance with SQL, and produced utilization reports.

JavaDockerSQLSystems support

03 / Jan 2025 – Present

H₀ ↔ H₁

Project Honeycrisp

RIT School of Interactive Games and Media · Rochester, NY

Quantitative analysis and research communication for instructional-media work.

Cleaned, integrated, and analyzed Project Honeycrisp data with R and MongoDB.
Produced visualizations and quantitative summaries to support research decisions.
Analyzed historical datasets, including A/B testing, and contributed figures, tables, and summaries for papers and presentations.

RMongoDBA/B testingData visualization

04 / Jan 2024 – May 2024

engagement ↑ 25%

Forge Innovation and Ventures

User-engagement analysis, KPI reporting, and product-strategy research.

Analyzed user-engagement data and recommended features that increased engagement by 25%.
Built concise KPI reports and partnered with cross-functional teams to streamline data workflows.
Researched emerging technologies and summarized product-strategy opportunities to support decision-making.

User engagementKPI reportsProduct strategy

05 / Jun 2023 – Aug 2023

O(n³) // GPU

National Atmospheric Research Laboratory · India

GPU-accelerated matrix operations and kernel benchmarking.

Developed CUDA-accelerated Gauss–Jordan routines for large matrix operations.
Benchmarked GPU kernels and documented performance insights for HPC workflows.
Integrated computational improvements within distributed environments in coordination with a networking team.

CUDAGPU kernelsGauss–JordanDistributed computing