I turn technical ideas into useful tools and experiences.
I'm Jay, a builder, writer, and speaker with roots in systems administration, cloud infrastructure, and developer advocacy. I learn by getting hands-on, asking questions, and sharing what I find. Whether I'm experimenting with AI, talking through lessons from an outage, or helping someone build their first app, I want to make technology easier to understand and give more people the confidence to create.
Technical product strategy AI-assisted development Developer experience
01 / Selected work
A fleet-operations sample where an agent reviews vehicle and charging data, then makes grounded recommendations. Includes an emulator-based CI workflow to test agent behavior with Cosmos DB.
A full-stack app for browsing movies, writing reviews, and viewing ratings and reviewer profiles. Runs locally against the Cosmos DB emulator.
A content research agent that scans public sources, filters and catalogs useful findings, and drafts reports and social posts. Use it from an editor or a browser dashboard.
Introducing SCOPE, an open-source approach to evaluating AI coding agents. A look at measuring agent capabilities beyond a generated code demo.
02 / Writing
Practical guides and ideas, newest first.
03 / Watch
Demos and conversations about building with AI and cloud databases.
04 / Talks, programs & engineering
AI evaluation / 2026
Built and ran coding-agent evaluations across multiple AI models and hundreds of generated samples using static analysis, compile-time checks, execution, and unit tests. Findings informed 20 merged Agent Kit pull requests and 27 public documentation improvements. In evaluated scenarios, a targeted update increased hierarchical partition-key adoption from 47% to 100% and reduced OFFSET/LIMIT usage from 47% to 0%. Co-presented From Findings to Fixes with Cedric Vidal and was featured in Microsoft’s SCOPE case study.
Agentic application development / 2026
Delivered Vibe Coding on Azure Cosmos DB: Build an AI App and Building APIs Faster with Agentic Development and Azure Cosmos DB. Demonstrated GitHub Copilot, Azure MCP, and the Azure Cosmos DB Agent Kit through application builds covering CRUD APIs, SDK integration, data modeling, partitioning, and indexing.
Conference / 2025
Co-presented Vibe coding: Ship it faster with GitHub Copilot and Azure Cosmos DB with Sajeetharan Sinnathurai. Demonstrated application development using GitHub Copilot, the Azure Cosmos DB Linux emulator, and the VS Code extension, giving developers a practical workflow for building and testing locally.
Technical discussion / June 2026
Joined Ravit Jain and Patty Chow for a technical discussion translating conference announcements into practical guidance for building AI applications. Covered agent memory, hybrid search, global scale, query cost visibility, and the Azure Cosmos DB Agent Kit.
Developer conference / 2023–2026
Led planning, production, and live hosting for the annual developer conference, including the five-hour 2026 broadcast. Built the event website, developed speakers and technical programming, and coordinated customer testimonials. The 2026 program covered AI agent memory, retrieval, data modeling, and production architecture.
Developer education / 2020–2022
Created and hosted more than 60 AzureFunBytes episodes, including walkthroughs of Bicep infrastructure deployment and Durable Functions automation. Produced and co-hosted MSCreate: DevOps, teaching automated Kubernetes deployment with Bicep, GitHub Actions, and Azure Key Vault.
Reliability / 2018–2020
Created and hosted the 48-episode On-Call Nightmares podcast, sharing incident-response and reliability lessons from practitioners including Kelsey Hightower, Silvia Botros, and Ryan Kitchens. Presented Checklists: Good Enough for NASA, Good Enough for Your App at DevOpsDays, translating operational experience into practical engineering habits.
Conference / 2018
Demonstrated Running Your First Docker Container in Azure, introducing container deployment through Azure Container Registry and Azure App Service. Published the accompanying sample application so developers could follow the workflow themselves.
Database deployment / 2017–2018
Published an Ansible playbook that provisioned MongoDB Atlas clusters through its API. Authored Modern Distributed Application Deployment with Kubernetes and MongoDB Atlas, demonstrating how to containerize a Node.js application with Docker, connect it to Atlas, and deploy it on Google Kubernetes Engine.
Technical training / 2017
Delivered Introduction to MongoDB, guiding developers from document database fundamentals through Atlas deployment, application connectivity, and data operations. Covered backups, alerts, and performance monitoring to connect initial adoption with operational readiness.
Production reliability / 2015; presented 2019
Served as the on-call engineer during the viral traffic surge that exceeded 670,000 active visitors. Worked with the engineering team to mitigate production failures, then shared lessons in incident response, scaling, and operational toil through What I Learned From a Dress, an On-Call Nightmare at DevOpsDays Portland.
05 / Keep in touch
