Website
Service
Full-Stack Engineering
AI Integration
Design Systems
A/B Testing
Technologies Used
Next.js, TypeScript, Styled Components, Builder.io, Amplitude, Segment, Okta, Framer Motion
As the most senior developer on the website team, I lead day-to-day technical decisions and mentor engineers to build a world-class digital experience for Productboard.
I developed and maintained a high-traffic website using Next.js, TypeScript, and Styled Components. To empower the marketing team, we utilized Builder.io as a headless CMS, allowing for marketer-friendly publishing at massive scale without developer bottlenecks. We also integrated complex interactive components using modern CSS and the Framer Motion SDK.
Working closely with the design system guild, I helped migrate the marketing site to unified design tokens using Token Studio. These updates are maintained and shipped efficiently through a custom internal npm package, ensuring consistency across all web properties.
To modernize our marketing operations, I led the technical migration away from legacy tools like Marketo to Conversion.AI and AirOps. Rather than relying on out-of-the-box embeds, I leveraged their APIs to build completely custom forms using our own internal React component library. Because both AirOps and Conversion.AI feature autonomous workflows, I architected a system that allows these tools to securely write content directly into our site.
Beyond third-party marketing tools, I built a direct integration with Spark, Productboard's own AI tool for product managers. I engineered a secure API layer using Okta authentication that empowers Spark to programmatically write and publish content to the marketing site, deeply intertwining our core product with our growth engine.
Allowing multiple AI systems to write directly to the site requires ironclad reliability. To guarantee 100% uptime, I engineered a highly resilient architecture around these APIs, introducing a "content mirror" and an automated fallback system. If any third-party service experiences downtime, our systems seamlessly fall back to the mirror, ensuring zero data loss. Applying this same philosophy of total resilience, I also architected a comprehensive backup and recovery system for the entire marketing website infrastructure.
To further support this AI-first ecosystem, I engineered custom Model Context Protocol (MCP) servers. These tools are deeply integrated into our workflows and are constantly used by the engineering team to automate complex development tasks directly through AI agents.
To drive data-informed decisions, I built an advanced component-level A/B testing framework directly into the marketing site, seamlessly integrated with Amplitude and Segment.
Throughout the development lifecycle, server usage and frontend performance are strictly tracked and optimized using the Vercel observability dashboard.