Projects

Featured work across compilers, AI infrastructure, and systems engineering.

These engagements blend deep compiler knowledge with platform pragmatism: Swift ↔ C++ interoperability, RISC-V vectorization wins, and resilient AI systems for high-stakes domains.

Keyda AI Tutoring Platform

Production RAG pipeline serving multiple school districts with >90% retrieval accuracy

RAGAWSpgvectorNext.jsMulti-tenant

Key achievements

  • Secure multi-tenant infrastructure with RBAC
  • Reduced LLM failure rates by 65%
  • Automated infrastructure via Terraform

DecorateAI GenAI Platform

Real-time visual processing inference pipeline with ARKit integration

GenAIComputer VisionARKitModel Optimization

Key achievements

  • Reduced latency by 40% through caching
  • Real-time model inference for interior design
  • Systematic prompt tuning framework

RISC-V Performance Optimization

Achieved 3x performance improvement on NASA benchmark suite at SiFive

LLVMClangRISC-VVectorizationCode Generation

Key achievements

  • Tuned vectorization cost models
  • Architecture-specific optimizations
  • Built CI/CD for cross-compilation

Swift-C++ Interoperability

Led interoperability initiative across Swift/Clang toolchain at Apple

SwiftC++ClangCompilerABI

Key achievements

  • AST bridging and name mangling
  • Coordinated Apple Silicon migration
  • Open-source contributions

Interested in working together?

Let’s design architectures that hold up under real-world stakes. I partner with teams shipping compilers, AI systems, and mission-critical infrastructure.

Start a conversation

Prefers engagements that blend performance engineering, tooling rigor, and measurable impact.