Engineering

Systems

Production.

Systems & Backend Engineer focused on turning ambiguous problems into production systems and high-leverage tools. Experienced in Go, Rust, distributed systems, AI-powered applications, and infrastructure, with end-to-end ownership across problem discovery, prototyping, deployment, and observability.

PRODUCTION EXPERIENCE

Work Experience.

Benchmark Administration

Golang Developer / Backend Engineer

Engineered and operated backend Go services in production with full ownership across service architecture, deployment pipelines, and operational observability.

  • Identified backend caching and state-synchronization requirements and designed a dedicated Redis service to improve application responsiveness and reduce repeated backend work. Owned the implementation, deployment, monitoring, and production operation end-to-end.
  • Reduced repeated database/backend operations by introducing a dedicated Redis caching layer, improving application responsiveness and creating a more scalable foundation for growing traffic.
  • Engineered backend Go services in production with full deployment ownership, health tracking, and operational monitoring.
  • Established end-to-end logging, metrics monitoring, and health tracking across production backend services.
Go (Golang)Redis ServiceBackend ArchitectureObservabilityDeployment Ownership
Open SourceContributions
iam-human-protocol

executor-node

PR #12 - Merged
Rust (Axum) · 2026
RustAxumAtomicU64AuthCaching
01

Implemented a live status endpoint exposing node uptime, verification and attestation counters, and relayer balance - giving operators real-time visibility into node health.

02

Designed a lock-free metrics layer using AtomicU64 for thread-safe, zero-contention counter updates across concurrent request handlers.

03

Implemented API key authentication with constant-time comparison (subtle::ConstantTimeEq) to prevent timing attacks.

04

Added a 30-second balance cache with TTL-based invalidation to avoid redundant RPC calls, alongside a full unit test suite.

SYSTEMS ARCHITECTURE & PROJECTS

Built for Scale & Low Latency.

Deck — AI Presentation Generator

Built an AI-powered presentation generation product using Gemini, with parallel slide generation and automated HTML/PDF/PPTX export. Designed the system around generation cost, latency, and scalability constraints to support high-volume presentation creation.

I/O PROTOCOLREST & Streaming LLM API
CONCURRENCYParallel Slide Generation Pipeline
KEY ENGINEERING DECISIONS
  • Designed parallel generation workflows to significantly reduce end-to-end presentation generation latency
  • Optimized prompt context and token pipeline to achieve cost efficiency (~$0.50 per presentation) while preserving output fidelity
  • Automated multi-format rendering pipelines supporting instant HTML preview, PDF export, and PPTX downloads
Gemini APIAI AutomationTypeScriptParallel GenerationMulti-Format Export

Aks Ide

A browser-based development environment designed to provide local-IDE-like terminal responsiveness while keeping execution isolated in containers.

I/O PROTOCOLWebSocket Binary Frame PTY Streaming
CONCURRENCYRust Async Tokio Task Pools
KEY ENGINEERING DECISIONS
  • Identified terminal latency and execution isolation as core constraints and designed a binary WebSocket/PTY architecture with containerized execution to address both
  • Architected custom PTY WebSocket stream to achieve sub-15ms terminal latency
  • Enforced container isolation with memory cgroups and ephemeral volume mount limits
RustContainerizationWebSocketsDockerWebAssembly

Video Search Engine

Built a semantic video search prototype to explore whether vector embeddings could provide more useful retrieval than traditional keyword search. Designed an event-driven ingestion pipeline using Kafka and Go/Python workers, with HNSW-based vector retrieval.

I/O PROTOCOLKafka Event Stream & REST
CONCURRENCYGo & Python Microservice Workers
KEY ENGINEERING DECISIONS
  • Hypothesized vector embedding retrieval outperforms keyword search for semantic video transcript discovery
  • Decoupled ingestion and embedding pipelines with Kafka to sustain high throughput during video indexing
KafkaPythonGoMicroservicesAmazon BedrockVector Search

KVDB - Database

A disk-backed, embedded key-value store in Rust - custom B+-tree storage engine with page-based persistence and a compile-time-enforced typestate API.

I/O PROTOCOLRust Typestate API
CONCURRENCYPage-level lock management & read-copy-update semantics
KEY ENGINEERING DECISIONS
  • Designed custom B+-tree storage engine with page-based disk persistence
  • Enforced compile-time transaction guarantees via Rust typestate pattern
RustB+TreeDisk storageConcurrency

Terminal Emulator

A GPU-accelerated terminal emulator written in Rust built for low-latency hardware-accelerated text rendering.

RustWgpuGUIGPU Acceleration

Akssora UI

An enterprise-grade UI component library designed for scalability and maintainability. Specialized in building accessible, theme-aware components with strict type safety. Engineered for tree-shaking and minimal bundle size impact.

Design SystemsMotionReact PerformanceTypeScript

Clipboard

A distributed clipboard synchronization system using Redis Pub/Sub for sub-millisecond data propagation across devices. Built with Go to handle thousands of concurrent connections.

I/O PROTOCOLWebSocket & Redis Pub/Sub
CONCURRENCYGo Channel Broadcast Fan-out
RedisGoConcurrencyDistributed Systems

What I Work With

Skills

PRODUCT & PROBLEM SOLVING

Product PrototypingTechnical DiscoveryAI AutomationExperimentationPerformance & Cost OptimizationDeveloper Tooling

BACKEND ENGINEERING & GO

Go (Golang)GinMicroservicesRust (Axum)PostgreSQLWebSocketsWebRTCgRPC

DevOps & OBSERVABILITY

DockerKubernetesKafkaRedisCI/CDLinux SystemsPrometheus Metrics

SYSTEM DESIGN

Distributed SystemsScalabilityHigh AvailabilityCAP TheoremLock-free Concurrency

CLIENT ARCHITECTURE

React jsNextjsTypeScriptPerformance ProfilingTailwind CSSFramer MotionGSAP

Systems &
Backend Engineer

Systems & Backend Engineer focused on turning ambiguous problems into production systems and high-leverage tools. Experienced in Go, Rust, distributed systems, AI-powered applications, and infrastructure, with end-to-end ownership across problem discovery, prototyping, deployment, and observability.

At Benchmark Administration, identified backend caching and state-synchronization requirements and designed a dedicated Redis service to improve application responsiveness and reduce repeated backend operations. Owned the implementation, deployment, monitoring, and production operation end-to-end.

Built AI-powered products like Deck with parallel slide generation and multi-format export, architected Aks IDE with containerized execution isolation and binary WebSocket/PTY streaming, and developed KVDB (a custom disk-backed B+-tree storage engine in Rust) alongside event-driven vector retrieval pipelines.