PROFESSIONAL WORK
Production AI systems, revenue-generating products, and secure LLM deployments at IBM (HashiCorp) and Moody's Analytics.
Projects and roles
Keysensus
Context-aware data-governance service at IBM (HashiCorp). Applies policy decisions using business and service context, instead of static field-only matching, across distributed production pipelines, with ML classification of sensitive data on AWS Bedrock and an auditable draft → review → publish → revoke lifecycle. The control layer that determines what data is safe for downstream AI and analytics systems to use.
GenAI Credit Memo Generation
Built credit-memo documentation generation on the Banking GenAI team using OpenAI and LangChain over Grid and Orbis data, saving analysts 8+ hours per report. Prompt engineering and Python for production: enterprise GenAI document workflows in regulated finance.
ETL Pipeline Optimization
Modernized high-throughput data pipelines at Moody's Analytics using Databricks and PySpark, achieving 54% reduction in processing time across critical workloads. Migrated legacy workflows to modern distributed computing architecture, the data-velocity foundation that downstream analytics and AI systems depend on.
Sanctions Compliance System
Built sanctions screening product at Moody's Analytics that generated $1.2M in first-year revenue. Automated compliance checks across financial transaction datasets with a 99.7% EU privacy compliance rate.
LangChain Citation System
Engineered LLM-powered document chunking and citation system at Moody's Analytics using LangChain and GPT-4 Turbo. Improved RAG citation accuracy by 92%, solving the hallucination problem and making AI-assisted research workflows trustworthy for financial analysts.
Data & Growth Consultant
Helped small businesses grow through SEO optimization, A/B testing, and Google Search Console analytics. Client work spanning local businesses and early-stage startups.
The Cookbook
AI-maintained engineering cookbook at IBM (HashiCorp). An LLM reads each thread in a busy internal help channel for a platform migration and distills it into a recipe: the problem, the fix, and the steps. Every new thread is checked against the existing recipes, so it either updates one or adds a new one, which keeps the documentation current on its own. Each recipe links back to its source thread and the engineer who answered it, so every answer is traceable and credited. Engineers new to the platform learn from the recipes instead of searching months of chat history, and answers given once get reused across the org.
The Path Here
Career arc, including the eight months between Moody's and IBM.
Impacted by Moody's layoffs
Posted publicly within the hour. Optimistic about what's next, grateful for the team I worked with.
Eight months of MSBA + side projects
Coursework at CMU Tepper, the Twitter Archive Pipeline, an Uma Musume / LangChain RAG refresh, a 48-hour game jam, a DIY home media backup, and an Artist Alley table mid-job-search.
Joined IBM (HashiCorp) as SRE II
Site Reliability Engineer II on the IOPS team. Contributed to Keysensus, the team's context-aware data-governance service, and shipped my production piece a few weeks in. The pipeline automation now saves 10 hours of manual review every week.