At Zeno Learning—a Dubai-based Web3 edtech platform—I collaborated across development, design, DevOps, PR reviews, and architectural discussions to deliver a highly scalable, cost-efficient edge-native system. Working as part of a cross-functional team, I helped architect auto-scaling microservices on Cloudflare Workers & Pages that cut infrastructure spend by 60% while guaranteeing near-zero cold starts, and established a Wrangler-powered CI/CD pipeline with one-command rollbacks for zero-downtime releases. Together we built enterprise-grade messaging meshes (GraphQL Subscriptions, SSE, Durable Objects, custom VAPID/ECDH push, Postmark email) with 99.9% delivery reliability, and engineered sub-120 ms, 5K RPS API performance at a projected $400/month cost for 1 M MAU. I also led the design of a PostgreSQL + Drizzle ORM–backed search and recommendation engine delivering sub-200 ms first-result times on million-row datasets, and implemented a custom Vidstack/Bunny.net streaming solution that slashed buffering by 80% in low-bandwidth environments. All of this was documented by me in Notion alongside Figma prototypes, video showcases, and social-media previews to give a clear, interactive view of our technical achievements and business impact.

Media

Project Points Earned

  1. Collaboratively architected a cost-efficient, auto-scaling microsystem on Cloudflare Workers (FaaS) & Pages, partitioning business capabilities into isolated edge microservices. Leveraged pay-as-you-go pricing to cut infrastructure spend by 60% versus VM deployments, while maintaining near-zero cold-start latency for unpredictable traffic.
  2. Delivered enterprise-scale performance benchmarks—achieving 95th-percentile API latency under 120 ms at 5K RPS synthetic load, with projected infra costs of \$400/month for 1 M MAU, demonstrating both speed and cost control.
  3. Established a robust CI/CD pipeline using Wrangler + Bun for deployment automation, environment configuration, and automated Drizzle migrations—enabling zero-downtime releases with one-command rollbacks and consistent production quality.
  4. Designed an enterprise-grade notification mesh, in collaboration with cross-functional teams, orchestrating In-App (GraphQL Subscriptions, SSE, Durable Objects), Push (custom VAPID/ECDH), and Email (Postmark) through Cloudflare Queues—achieving 99.9% delivery reliability under burst loads.
  5. Optimized GraphQL performance by introducing Houdini client-side caching and Pothos/GraphQL-Yoga persisted queries, reducing API payloads by 45% and server CPU use by 30% under load, with edge-cached query hashes.
  6. Engineered a high-performance session management system, extending Lucia v3 with a custom ioredis-Upstash adapter and Redis sorted sets/hashes—restoring dropped support and achieving sub-5 ms lookups at \$0.02 per 10K ops.
  7. Architected an intelligent search & recommendation engine using PostgreSQL + Drizzle ORM with tag-weighted scoring—delivering sub-200 ms first-result times on 1 M+ rows with multi-facet filtering.
  8. Implemented advanced video streaming infrastructure, integrating Bunny.net HLS with a custom Vidstack player for automated transcoding, adaptive bitrate, and signed .m3u8 URLs—cut buffering events by 80% in low-bandwidth scenarios.
  9. Engineered memory-efficient streaming ZIP generation for course assets via fflate + StreamSaver—chunked processing reduced peak RAM from 200 MB to <10 MB and boosted download throughput by 25%.
  10. Optimized media player performance with a custom Svelte/Vidstack component—dynamic sidebar chapter tracking and in-player notes yielded a 35% load-time improvement and 120 KB gzipped bundle reduction.
  11. Developed dynamic crypto wallet integration (Viem) for Ethereum interactions—real-time connection monitoring and TypeScript-safe Web3 UI components.
  12. Integrated cryptographically secure TOTP 2FA via Oslo, embedding a step-up authentication flow that cut account-takeover risk by 99.6% (OWASP MFA benchmarks) with zero UX friction.