Where I've worked

IBM Verify · Remote

Jun 2026 – Sep 2026

Built agentic AI systems for enterprise identity and access management, combining LLMs with automated reasoning to automate identity risk analysis.

• Built a multi-agent identity risk workflow in Langflow routing cases across 4 specialized agents behind Python services and APIs, cutting end-to-end analysis from ~20 minutes of manual review to under 2 minutes per case.
• Designed an MCP-based tool-filtering layer that scopes an agent's available actions to a user's entitlements, cutting exposed tool surface by over 15% across 5 workflows.
• Traced ±8% run-to-run behavior variance to tool-ordering nondeterminism and reduced it to ±3%.

• Python
• LLMs
• Automated Reasoning
• Langflow
• MCP
• APIs

Inspirit AI

Jun 2026 – Present

• Teach Python, machine learning, and AI concepts through project-based instruction, leading small groups of 5-6 students per cohort.
• Mentor students in designing and implementing end-to-end AI projects applied to fields such as education and investment, using real-world datasets.
• Provide technical guidance on model development, evaluation, and optimization.

• Python
• Machine Learning
• AI

Develop for Good · Stanford, CA

Oct 2025 – Feb 2026

Led an accessible, cognitive-friendly website redesign for ABI patients and caregivers, shipped across 8 pages and flows.

• Led an end-to-end redesign shipped across 8 pages and flows as PM with Develop for Good, incorporating feedback from 10+ user-testing participants.
• Defined product requirements, user flows, and feature scope, iterating with nonprofit stakeholders.
• Coordinated design and low/no-code implementation across a team of engineers and designers.
• Partnered with Brain Injury Connection to deliver an intuitive resource platform aligned with its mission and user needs.

Gordian Knot Center for National Security Innovation · Stanford, CA

Sep 2025 – Jun 2026

Founded and led a biosurveillance "digital immune system" research project.

• Fused 3 independent data sources into a single risk index, validated against 50+ synthetic scenarios spanning baseline, outbreak, and stress-test conditions.
• Selected an interpretable linear fusion model over higher-capacity alternatives to preserve explainability for decision-makers.
• Defined precision, recall, and latency thresholds; achieved 88% precision / 82% recall at 150ms median latency.
• Deployed a secure, fault-tolerant data infrastructure with containerized services, API gateways, and automated recovery for resilient biosecurity signal processing.

• Python
• SQL
• Time-Series Modeling
• ETL
• Containerization

Medtronic Diabetes · Los Angeles, CA

Jun 2025 – Aug 2025

Automated test workflows and analytics for a next-generation blood-glucose analyzer.

• Automated test data collection and visualization with Python and REST APIs for a new glucose analyzer, cutting manual data-processing time by 45%.
• Rebuilt the test-data ETL path with secure pipelines and Git-based versioning, cutting runtime from ~25 minutes to ~6 minutes per batch.
• Benchmarked device performance with statsmodels and scikit-learn across 200+ test runs for statistical and regression analysis.

• Python
• REST APIs
• statsmodels
• scikit-learn
• Minitab
• Git

Stanford Marsden Cardiovascular Biomechanics Lab · Stanford, CA

Feb 2025 – Jun 2026

Built a patient-specific computational pipeline evaluating virtual stenting in the Fontan circulation, as sole student researcher.

• Built a patient-specific pipeline from 4D flow MRI to CFD as sole student researcher: reconstructed 15 Fontan anatomies, screened to 3, generated paired virtual-stent geometries in svMorph, and ran 12 transient simulations at 32,000 time steps each.
• Tuned branch-specific RCR Windkessel outlet conditions until simulated flow splits and mean pressures matched catheterization and MRI data; simulated moderate exercise (3× IVC inflow, 120 bpm) alongside rest.
• Virtual stenting cut pressure drop 34.4–97.9% and resistance 70.4–99.5% across all three patients, but power loss rose 45.6% in one under exercise, showing a wider pathway is not automatically a more efficient one.

• SimVascular
• svMorph
• Python
• Computational Fluid Dynamics

Stanford Medicine — Medical Kiosk · Stanford, CA

Jan 2025 – Present

Kiosk software to expand healthcare access in underserved communities.

• Build multilingual intake modules on REST APIs and SQL backends, covering 4 languages across 3 intake flows.
• Refined intake UX through feedback loops and A/B testing, cutting mid-intake abandonment from 22% to 9%.
• Integrate medical-device APIs behind Dockerized microservices for scalable deployment and cross-device synchronization.

• Python
• REST APIs
• SQL
• Docker
• Microservices

Stanford University · Stanford, CA

Sep 2024 – Present

• Foster a positive, inclusive, and supportive residential community by building meaningful connections with residents.
• Plan community programming and serve as a first point of contact for student support and resources.

Stanford Department of Medicine · Stanford, CA

Jun 2024 – Aug 2024

• Mentored high-school scholars in a Stanford lab on hands-on biomedical research.
• Taught core techniques (histology, imaging, immunohistochemistry) and coached experimental design and data analysis.
• Guided students to a final presentation showcase while fostering an inclusive, curiosity-driven environment.

Stanford Alternative Spring Break · Stanford, CA

May 2024 – Jun 2025

Elected Financial Officer of Stanford's second-largest Volunteer Service Organization (10 trips, ~100 participants, 20 co-leads).

• Oversaw budgeting, fundraising, and financial sustainability across the organization's 10 service trips.
• Co-led the Affordable Lifeline trip to Alaska, designing the curriculum and program to explore healthcare access in remote environments.
• Managed finances and logistics and facilitated hands-on service projects examining rural healthcare challenges.

Stanford University School of Medicine · Stanford, CA

Apr 2024 – Present

• TA for BIOE 122, “Biosecurity and Pandemic Preparedness.”
• Support instruction by facilitating discussions, guiding student projects, and providing feedback.

Stanford Computer Science Department · Stanford, CA

Sep 2023 – Jun 2026

• Mentored 10+ students per quarter across 6+ quarters in C++, Python, and JavaScript, teaching code design and debugging practices.
• Debugged and reviewed 100+ code submissions quarterly, catching recurring design issues and promoting version control and testing methodology.

• C++
• Python
• JavaScript

Stanford Fallacy Detection CS Research Project · Stanford, CA

Jun 2023 – Dec 2023

Benchmarked large language models for logical-fallacy detection.

• Benchmarked LLMs for fallacy detection and built baselines in Python (transformers, scikit-learn) with precision/recall/F1 reporting.
• Curated datasets and prompts; implemented ETL, feature engineering, and reproducible evaluation pipelines.
• Co-developed a fallacy detector and ran ablation studies and error analysis to refine model design.

• Python
• Transformers
• scikit-learn