Open Source Systems & Utilities•Native macOS

Open Source Projects & Architecture Notes

Open-source native applications, high-performance systems engineering, and architectural notes focused on privacy-first computation and zero-knowledge guarantees.

Akshay Kr Gupta

Senior Solutions Architect & Founder

“Designing enterprise systems for regulated environments — where architecture, AI, and governance intersect. Open to interesting architecture conversations & collaborations.”

Personal Website•Architecture Notes•GitHub•LinkedIn

Technical Architecture Notes

Essays on enterprise systems design, AI governance, and distributed computing.

Read Architecture Notes

In-depth technical publications covering enterprise architecture in regulated environments, Model Context Protocol (MCP) integrations, high-throughput microservices, and AI compliance governance.

Enterprise ArchitectureModel Context Protocol (MCP)Distributed SystemsAI GovernanceCloud Native & Kubernetes

Featured Native Projects

Engineered natively with Swift, SwiftUI, and Apple Frameworks for 100% offline privacy.

TwinPixCleaner

v2.2.0Apache 2.0

AI-Powered Visual Duplicate & Similar Photo Cleaner for macOS

Releases View Repository

A native macOS utility built with SwiftUI and Apple Vision Intelligence that accurately detects duplicate photos, visually similar shots, burst series, and edited screenshots without uploading a single pixel to the cloud.

Technology Stack & Frameworks

Swift 6.0SwiftUIApple Vision Framework (Rev2)Accelerate SIMD (vDSP)PhotoKitAppKit

Key Features & Capabilities

•Dual Scanning Engines: Pure SHA-256 binary hash matching and Apple Vision ML feature prints.
•Transitive Union-Find Clustering powered by Accelerate SIMD vector distance math (vDSP_distancesq).
•Native Apple Photos integration via PhotoKit to securely inspect your entire iCloud library.
•Persistent on-disk similarity cache for lightning-fast incremental subsequent scans.
•Interactive Skipped Files Inspector sheet with search, categorizations, and clipboard export.
•Safe non-destructive deletion with macOS Trash / Recently Deleted integration and ⌘Z instant Undo.
•Native Quick Look (Spacebar) preview, keyboard shortcuts, and full VoiceOver accessibility.

System Architecture Highlights

High-performance ML feature extraction running locally on Apple Neural Engine / GPU.
Zero-allocation SIMD vector math for clustering thousands of photo embeddings in milliseconds.
Zero telemetry, zero external network calls — 100% offline privacy guarantee.

Build from source

# Clone & build native release
git clone https://github.com/AkshayKrGupta/TwinPixCleaner.git
cd TwinPixCleaner
swift build -c release
./.build/release/TwinPixCleaner

NanoPress

v1.0.0Apache 2.0

Private, Local-Only Parallel Batch Compressor for Images & PDFs

Releases View Repository

A lightning-fast native macOS utility built with SwiftUI and Swift Concurrency to compress JPG, PNG, HEIC, TIFF images, and PDF documents locally on your device with zero cloud uploads.

Technology Stack & Frameworks

Swift 5.9+SwiftUISwift Concurrency (Actors, TaskGroup)CoreGraphicsImageIOPDFKit

Key Features & Capabilities

•100% Local Processing: Zero cloud uploads, zero tracking, and zero latency.
•Multi-Format Image Compression: Supports JPG, PNG, HEIC, and TIFF with custom quality sliders.
•Dual-Engine PDF Optimization: Standard mode for sharp quality and Aggressive mode for maximum size reduction.
•High-Throughput Parallel Batch Engine: Uses Swift Actors and Task Groups to process hundreds of files concurrently.
•Live Size Savings Preview: Inspect before-and-after byte reductions before committing.
•Drag-and-Drop Workflow: Seamless drag-and-drop ingestion with adaptive Light and Dark mode.

System Architecture Highlights

Actor-isolated pipeline preventing thread race conditions during heavy multi-core batch compression.
Direct CoreGraphics and ImageIO pipeline for memory-efficient streaming without buffering entire archives.
Native Swift Package Manager architecture with clean MVVM separation.

Build from source

# Clone & open in Xcode
git clone https://github.com/AkshayKrGupta/NanoPress.git
cd NanoPress
open Package.swift

Editorial Disclaimer & Copyright

The technical analyses, design patterns, and opinions expressed in this publication are solely my own and do not represent the views, positions, or strategies of my employer or clients.

Back to Tools

Open Source Systems & Utilities•Native macOS

Open Source Projects & Architecture Notes

Open-source native applications, high-performance systems engineering, and architectural notes focused on privacy-first computation and zero-knowledge guarantees.

Akshay Kr Gupta

Senior Solutions Architect & Founder

“Designing enterprise systems for regulated environments — where architecture, AI, and governance intersect. Open to interesting architecture conversations & collaborations.”

Personal Website•Architecture Notes•GitHub•LinkedIn

Technical Architecture Notes

Essays on enterprise systems design, AI governance, and distributed computing.

Read Architecture Notes

In-depth technical publications covering enterprise architecture in regulated environments, Model Context Protocol (MCP) integrations, high-throughput microservices, and AI compliance governance.

Enterprise ArchitectureModel Context Protocol (MCP)Distributed SystemsAI GovernanceCloud Native & Kubernetes

Featured Native Projects

Engineered natively with Swift, SwiftUI, and Apple Frameworks for 100% offline privacy.

TwinPixCleaner

v2.2.0Apache 2.0

AI-Powered Visual Duplicate & Similar Photo Cleaner for macOS

Releases View Repository

Technology Stack & Frameworks

Swift 6.0SwiftUIApple Vision Framework (Rev2)Accelerate SIMD (vDSP)PhotoKitAppKit

Key Features & Capabilities

•Dual Scanning Engines: Pure SHA-256 binary hash matching and Apple Vision ML feature prints.
•Transitive Union-Find Clustering powered by Accelerate SIMD vector distance math (vDSP_distancesq).
•Native Apple Photos integration via PhotoKit to securely inspect your entire iCloud library.
•Persistent on-disk similarity cache for lightning-fast incremental subsequent scans.
•Interactive Skipped Files Inspector sheet with search, categorizations, and clipboard export.
•Safe non-destructive deletion with macOS Trash / Recently Deleted integration and ⌘Z instant Undo.
•Native Quick Look (Spacebar) preview, keyboard shortcuts, and full VoiceOver accessibility.

System Architecture Highlights

High-performance ML feature extraction running locally on Apple Neural Engine / GPU.
Zero-allocation SIMD vector math for clustering thousands of photo embeddings in milliseconds.
Zero telemetry, zero external network calls — 100% offline privacy guarantee.

Build from source

# Clone & build native release
git clone https://github.com/AkshayKrGupta/TwinPixCleaner.git
cd TwinPixCleaner
swift build -c release
./.build/release/TwinPixCleaner

NanoPress

v1.0.0Apache 2.0

Private, Local-Only Parallel Batch Compressor for Images & PDFs

Releases View Repository

A lightning-fast native macOS utility built with SwiftUI and Swift Concurrency to compress JPG, PNG, HEIC, TIFF images, and PDF documents locally on your device with zero cloud uploads.

Technology Stack & Frameworks

Swift 5.9+SwiftUISwift Concurrency (Actors, TaskGroup)CoreGraphicsImageIOPDFKit

Key Features & Capabilities

•100% Local Processing: Zero cloud uploads, zero tracking, and zero latency.
•Multi-Format Image Compression: Supports JPG, PNG, HEIC, and TIFF with custom quality sliders.
•Dual-Engine PDF Optimization: Standard mode for sharp quality and Aggressive mode for maximum size reduction.
•High-Throughput Parallel Batch Engine: Uses Swift Actors and Task Groups to process hundreds of files concurrently.
•Live Size Savings Preview: Inspect before-and-after byte reductions before committing.
•Drag-and-Drop Workflow: Seamless drag-and-drop ingestion with adaptive Light and Dark mode.

System Architecture Highlights

Actor-isolated pipeline preventing thread race conditions during heavy multi-core batch compression.
Direct CoreGraphics and ImageIO pipeline for memory-efficient streaming without buffering entire archives.
Native Swift Package Manager architecture with clean MVVM separation.

Build from source

# Clone & open in Xcode
git clone https://github.com/AkshayKrGupta/NanoPress.git
cd NanoPress
open Package.swift

Editorial Disclaimer & Copyright

The technical analyses, design patterns, and opinions expressed in this publication are solely my own and do not represent the views, positions, or strategies of my employer or clients.