About
Ilteris Kaplan
Senior Design Technologist | AI Infrastructure & Prototyping
I build the foundational systems that make entire teams faster.
Design Technologist with 11 years at Google delivering zero-to-one product visions for Cloud AI. I architect the infrastructure, prototypes, and developer tools that transform ambiguous AI product concepts into production-ready experiences.
At Google Cloud AI, I led keynote demos for Cloud Next '25 and '26, built a 9-pass agentic pipeline that achieved 48x analysis speedup, and spearheaded the GitHub Enterprise platform that 200+ people use daily. Before Cloud AI, I spent 5 years on Google Jamboard & Meet building for collaborative hardware, and before Google I shipped 6 iOS apps to the App Store and built interactive installations for brands like Moet Chandon and Hennessy.
Core Principles
- From Autocomplete to Agency: I design systems where AI isn't just suggesting code, but executing and self-correcting within clean local sandboxes and shell primitives.
- Platform-Level Velocity: I believe prototyping requires robust infrastructure. I build the CLIs, template engines, and deployment pipelines that reduce setup friction from days to minutes.
- Managed Interception: Autonomous agents need boundaries. I build structured hook systems and verification gates that keep execution aligned, preventing drift while maintaining autonomy.
This blog is a living archive of 20+ years of technical work — from early experiments in physical computing at NYU ITP to modern agentic systems and AI infrastructure.
Career Arc
Google Cloud AI
Senior UX Engineer & Design Technologist | 2020 – 2026
I architected the prototyping infrastructure, developer platforms, and agentic workflows across Vertex AI and Gemini Enterprise to de-risk zero-to-one product investments:
- Keynote Prototypes: Lead developer of the conversational commerce shopping engine showcased at Cloud Next '25 (later launched as a GA Vertex AI product) and the partner-agent procurement flow for Cloud Next '26.
- Agentic UXR Platform: Designed a dual-service agentic pipeline (FastAPI, Google ADK, Vue 3, Firestore) that automated qualitative research analysis across 9 sequential passes (including adversarial quality judging and human-in-the-loop checkpoints), reducing study analysis time from roughly 4 hours to 5 minutes (48x speedup).
- Developer Platforms: Spearheaded the adoption of GitHub Enterprise for Google Cloud UXE. Authored the migration roadmap and developed the
vibeCLI toolchain, enabling 50+ UX/PMs/Engineers and multiple teams across Google to deploy prototypes 5-10x daily. - Figma MCP Server: Built an open-source Python Model Context Protocol (MCP) server exposing Figma's API to agents, handling node token limits and enabling automated component-to-code pipelines.
Google Jamboard & Meet
UX Engineer | 2015 – 2020
Founding UX Engineer on Google Jamboard—Google's 4K collaborative touch hardware—contributing to cross-platform experiences (web, mobile, and kiosk display) from initial concepts through its global 2018 General Availability launch.
Freelance & Interactive Media
Design Technologist | 2007 – 2015
Shipped 6 native iOS apps to the App Store, built e-commerce systems (generating a 150% revenue lift for TSE Cashmere), and engineered interactive physical and computer-vision installations for luxury brands including Hennessy and Moët & Chandon.