Kevin Szeto
@kevinszeto
Senior AI engineer building scalable RAG and production systems that power customer-facing products.
United StatesWhat I'm looking for
I’m a senior AI full-stack engineer with 10+ years of experience building production systems across e-commerce and enterprise SaaS. My focus is deep backend work in Python, Java, and Go, paired with applied AI—LLM integration, RAG pipelines, and vector search—to deliver measurable improvements for real users.
At Atlassian, I built the document ingestion and chunking pipeline for Rovo using Python and LangChain, reducing average query latency by 38% while keeping retrieval accuracy above 91%. I designed RAG architecture with OpenAI embeddings and Pinecone, scaled the search backend to 50M+ embedding lookups/day, and cut p99 latency from 1.4s to 340ms—also reducing cost per query by 29% via semantic caching with Redis. I’ve also led cross-functional work on data freshness SLAs with incremental re-indexing in Go, shipped collaborative editing using WebSockets, and mentored engineers through weekly reviews and internal documentation that standardized AI-integration practices.
Experience
Work history, roles, and key accomplishments
Built Rovo document ingestion and RAG pipeline, reducing average query latency by 38% while maintaining >91% retrieval accuracy. Scaled the search backend to 50M+ embedding lookups/day, cutting p99 latency from 1.4s to 340ms, and improved answer relevance by 44% in A/B tests.
Rebuilt Confluence page tree rendering in React, cutting initial render time by 52% for spaces with 10,000+ pages. Implemented high-throughput automation and observability improvements, reducing MT TTI from 47 minutes to under 12 minutes and enabling 2.3x throughput via Go microservices on Kubernetes.
Built core storefront and platform features in Ruby on Rails and JavaScript, supporting growth from under 200,000 merchants to 500,000+ active stores. Developed the Storefront API and Discount rules engine, and improved checkout reliability by reducing address-related errors by 31% and increasing completion rate by 8%.
Education
Degrees, certifications, and relevant coursework
University of California, San Diego
Bachelor of Science, Computer Science
2008 - 2012
Earned a B.S. in Computer Science from the University of California, San Diego from 2008 to 2012.
Tech stack
Software and tools used professionally
Availability
Location
United StatesAuthorized to work in
Job categories
Skills
Interested in hiring Kevin?
You can contact Kevin and 90k+ other talented remote workers on Himalayas.
Message KevinFind your dream job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
