pgtriage Assistant: an AI chatbot built with design thinking
A live AI assistant that helps engineers triage PostgreSQL performance problems, designed, built, guardrailed, and tested end to end as part of the AIML‑500 AI Lab.
Backend engineer at a fintech clearing platform processing over a million transactions an hour. I work where production reliability, database internals, and practical AI tooling meet.
-- before Seq Scan on account_transactions Filter: ((account_id)::numeric = $1) Rows Removed by Filter: 39,998,507 Execution Time: 45,210 ms -- fix: bind the right type, index for the query CREATE INDEX CONCURRENTLY ON account_transactions (account_id, created_at); -- after Index Scan using idx_account_created Execution Time: 23 ms
As a senior software engineer specializing in distributed systems and PostgreSQL performance in financial services, I help engineering organizations keep mission‑critical platforms fast, reliable, and auditable at scale. I pair deep production incident experience with modern AI tooling, turning hard‑won operational expertise into tools and assistants that engineering, operations, and compliance teams can use without me in the room.
Who this portfolio is for: hiring managers and technical leaders evaluating engineers for senior backend and forward‑deployed engineering roles. Each artifact demonstrates a competency those roles depend on: building AI products, communicating technical research, judging which AI approach fits a problem, and leading its adoption responsibly.
I am a Senior Software Engineer at Apex Fintech Solutions in Austin, Texas, where I own platform infrastructure on the Accounts Platform, the system of record for trading, funding, and compliance workflows across tens of millions of brokerage accounts. My day‑to‑day spans Java, Go, and Python services, PostgreSQL at multi‑million‑row scale, Apache Kafka event pipelines, and Kubernetes on Google Cloud.
The work I find most satisfying sits at the sharp end of production: leading P0 incident investigations across team boundaries, diagnosing query plans that silently defeat their indexes, and redesigning event architectures so the next migration does not become the next incident. That experience convinced me that the best engineering knowledge is worthless if it stays in one person's head, which is why I build AI tooling that codifies it, including an MCP‑based accounts platform adopted by three teams and pgtriage, an open‑source PostgreSQL performance auditing server.
I hold a Master of Science in Computer Science from Texas State University and am currently deepening my machine learning foundations in graduate coursework at Indiana Wesleyan University. I write about database internals and incident debugging on Medium.
A live AI assistant that helps engineers triage PostgreSQL performance problems, designed, built, guardrailed, and tested end to end as part of the AIML‑500 AI Lab.
A seven‑slide research presentation arguing that every major AI leap, from the artificial neuron in 1943 to today's large language models, was unlocked by compute, data, and distributed systems as much as by algorithms.
A comparative analysis of two mirror‑image production cases, churn prediction and machine translation, arguing when traditional machine learning beats deep learning and when it structurally cannot compete.
A leadership value statement pairing every fairness principle with an operational practice, from override capture and label audits to postmortems that name the hypothesis the team anchored on.
A subscriber‑ready newsletter analyzing how new AI products are reshaping healthcare, education, and news‑media workflows while preserving human authority over consequential decisions.