🎓 CS & Applied Math @ NYU • AI/ML

CI/CD release gate for LLM applications: catch PII leaks, prompt injection, and quality regressions before production.

Repo

• Built CLI with 6 evaluators: exact-match, contains, llm-judge, PII detection, prompt injection, custom
• Multi-layer prompt injection detection (heuristic patterns + canary tokens + LLM classification)
• PII scanner with Luhn validation for credit cards, SSN format validation, GDPR-relevant entity types
• Supports OpenAI, Anthropic, Azure, and custom OpenAI-compatible endpoints
• 159 unit tests, verified with real LLM API calls

Tech: TypeScript, Python, Node.js, Zod, Vitest

Predicting loan default with production-realistic constraints.

Repo

• Trained and evaluated models on 1.3M+ LendingClub loans with strict leakage prevention
• Addressed class imbalance and benchmarked logistic regression, random forest, and gradient boosting
• Translated probabilities into decision thresholds (approve / review / reject) rather than raw scores

Tech: Python, pandas, NumPy, scikit-learn

Business question → structured analysis (JSON, spreadsheet, deck).

Repo

• Converts vague prompts into explicit analysis plans using agent graphs
• Retrieval-grounded generation to ensure outputs are traceable to source data
• Benchmarked cost and latency tradeoffs across routing strategies

Tech: LangGraph, FastAPI, Redis, Docker

Framework to distinguish real trading signal from noise.

Repo · Demo

• Deterministic backtests with transaction cost modeling
• Statistical validation via bootstrap CI, permutation tests, and Monte Carlo simulation
• Explicit verdicts: edge, inconclusive, or noise — not just Sharpe ratios

Tech: Python, pandas, NumPy, SciPy, Streamlit, pytest

NLP + hypothesis testing on lyrical evolution.

Repo · Demo

• Sentiment analysis, lexical diversity metrics, topic modeling
• Statistical tests across albums (not just visualization or anecdotes)

Tech: Python, scikit-learn, React, Vite

Languages: Python, SQL, Java, TypeScript, C++
Data Engineering: PostgreSQL, AWS (S3, Redshift), Spark, Kafka, Airflow, Prefect, Docker, dbt
Analytics & ML: pandas, NumPy, scikit-learn, XGBoost, Tableau, Streamlit
AI/LLMs: LangChain, LangGraph, RAG pipelines, pgvector, multi-agent systems
Core Competency: ETL pipelines, data modeling, feature engineering, statistical analysis

💡 I like building things that actually work, then figuring out why they work.