π Hey, I'm David
I build production-grade systems that sit at the intersection of data engineering, backend platforms, and product.
Most days you'll find me:
- Designing scalable data pipelines (Spark, Airflow, Snowflake, Kafka, Databricks)
- Shipping backend services in Kotlin / Spring Boot
- Building cross-platform products with Flutter and Next.js
- Wiring up payment flows, multi-currency pricing, and marketplace infrastructure
- Coordinating multi-repo systems (backend β cloud functions β mobile β web)
- Obsessing over clean abstractions, idempotency, and systems that don't page you at 3am
Currently building: An experience marketplace platform end-to-end β Kotlin/Spring Boot APIs(hosted on Cloud Run), Firebase Cloud Functions, Data Connect, Flutter mobile app, and a Next.js web client β with real-time bookings, ticketing, and international payment processing.
Infra - Terraform, GCP
AI-Native Development
I write code with Cursor, Claude Code, and OpenAI Codex as core parts of my workflow β not as autocomplete, but as genuine force multipliers:
- Cursor as my primary IDE β inline AI assistance for rapid iteration across Kotlin, Dart, TypeScript, and React
- Claude Code as a pair programmer for complex, multi-repo tasks β I architect the plan, then execute full-stack features spanning 4 codebases in a single session (backend β cloud functions β mobile β web)
- OpenAI Codex for autonomous task execution and parallel workstreams
- AI-assisted cross-codebase debugging β tracing issues across service boundaries (e.g. tracking why a Firestore document isn't created from Spring Boot through Cloud Functions to the mobile client)
- AI-driven merge conflict resolution on complex domain logic files
- Comprehensive PR generation with structured descriptions across multiple repositories
I treat AI tooling the way I treat any good infrastructure: learn its strengths, set clear boundaries, and let it handle the throughput while I focus on the architecture and product decisions.
I care deeply about robustness, clarity, and shipping things that actually get used.