Enterprise Platforms
Modernization, API contracts, cloud architecture, integration governance, security patterns, and production-ready operating models.
peyi.dev / Personal CV
Principal Software Architect building enterprise platforms, AI-enabled engineering systems, and high-transaction commerce architectures.

I turn ambiguous business and technical needs into durable platform systems: architecture standards, service contracts, automation, integration strategy, delivery alignment, and production readiness.
Architecture Focus
Modernization, API contracts, cloud architecture, integration governance, security patterns, and production-ready operating models.
Adobe Commerce, Magento, pricing, inventory, checkout, loyalty, customer identity, content synchronization, and transactional consistency.
RAG, LLM evaluation, hallucination reduction, semantic classification, claim-level verification, and AI quality gates.
Architecture governance, distributed teams, delivery risk, stakeholder alignment, engineering standards, and launch readiness.
Selected Systems
Cisco / Splunk
Shapes technical direction for product experience platforms, content operations, automation, AI-enabled workflows, developer productivity, and platform modernization.
GoToFoods / Focus Brands
Architects restaurant technology across Olo, Contentful, Punchh, customer identity, mobile app integrations, checkout flows, Azure services, and operational automation.
Central Group
Led architecture strategy for a major retail ecosystem spanning multi-brand commerce, platform consolidation, performance, scalability, and governance.
Adobe Commerce Ecosystem
Designed enterprise commerce architectures for major retail, telecom, logistics, consumer product, restaurant technology, and marketplace programs.
Research
Papers, pilots, and technical notes.
Distributed systems / Commerce
Formal paper with reproducibility artifact
Contract limits, escrow safety, and risk-aware rights allocation for ATS reservation systems. The artifact includes simulator code, checksums, tests, and 60,000 run-level records across AP, CP, and escrow models.
01AI reliability / RAG evaluation
Hallucination-reduction pilot for retrieval-augmented AI
An AI reliability study for Retrieval-Augmented Generation (RAG): a paired evaluation of an unguarded assistant against a guardrailed retrieval-and-abstention bundle across 91 eligible production requests. The pilot targets hallucination reduction, unsupported-claim detection, abstention behavior, evaluator validity, and token/latency tradeoffs.
02Project Gallery
Architecture case
Research-to-platform pattern
Portfolio infrastructure
Toolbox
Education and research
Doctoral research focus: AI reliability, retrieval-augmented generation, hallucination reduction, claim-level verification, and quality systems for technical platforms.