Senior Backend / Platform Engineer building fintech and AI systems

Senior Engineer at Franklin Templeton · San Ramon, California · 10+ years

LinkedIn · GitHub · Medium · Email

I build backend and platform systems where correctness, throughput, and observability matter: transaction processing, reconciliation workflows, ETL and data platforms, and service integrations that have to hold up under production load.

My recent public work also leans into applied AI: RAG search, evaluation-aware LLM workflows, and developer-facing products that combine structured data, document pipelines, and model-driven features.

Most of my day-to-day production work lives in private fintech repositories, so the public projects below are representative samples of how I design systems, not the full body of shipped work.

• Prevented roughly $1M in incorrect payments by fixing payout calculation and settlement bugs in a large-scale payments flow at Angi.
• Reduced database load by ~40% by introducing distributed caching for high-throughput automotive retail services at Cox Automotive.
• Cut duplicate professional profiles by 15% with PII-aware detection and merge workflows that improved trust and data quality.
• Reduced page-load issues by 30% while leading feature delivery and performance fixes for a multi-team home services platform.

AI-powered resume builder that parses resumes, analyzes job descriptions, and generates tailored PDF and DOCX outputs with application-support workflows.

Built with Next.js 16, TypeScript, Prisma, document processing pipelines, and AI orchestration modules around profile and job analysis.

Strong evidence of end-to-end product ownership, backend API design, integration-heavy workflows, and applied AI shipped as a usable product.

Conversational property search platform that combines structured filters with RAG over a pgvector-backed real-estate dataset.

Built with Django, React, PostgreSQL with pgvector, OpenAI models, and analytics around chat quality and retrieval relevance.

Shows how I approach multi-service systems, search architecture, embeddings, and production-minded AI features beyond prototype demos.

Facial emotion recognition project comparing multiple ML approaches and exposing the results through a FastAPI web app with live inference.

Built with Python, FastAPI, TensorFlow, transfer learning pipelines, evaluation scripts, and a lightweight web frontend.

Useful proof of hands-on model experimentation, inference API design, and the engineering work required to turn ML experiments into an interactive system.

• Backend and distributed systems: Python, Go, Java, microservices, REST and gRPC APIs, fault-tolerant processing, and service integrations.
• Fintech data and reliability: transaction flows, reconciliation, idempotent pipelines, data integrity, ETL, Snowflake, and Airflow.
• Applied AI systems: RAG, embeddings, prompt and workflow design, evaluation-minded AI products, agentic workflows, and developer tooling.
• Cloud and platform engineering: AWS, Docker, Kubernetes, CI/CD, observability, and pragmatic production operations.

I write occasionally on Medium about data engineering, cloud architecture, and applied AI systems. I use GitHub mostly to publish representative projects, document technical decisions clearly, and keep public work easy to evaluate.

Outside of engineering, I am a serious car enthusiast. I have owned 5+ cars over the years, and I have a collection of 1,000+ toy cars.

If you are hiring for backend, platform, fintech infrastructure, or applied AI engineering roles, reach me at [email protected] or on LinkedIn.