Full Stack Developer | Go Backend Expert | Angular & TypeScript Specialist

I'm a passionate full-stack developer with deep expertise in modern web technologies and systems programming. I believe in building high-quality, production-ready open-source software that solves real problems.

Backend

• Go 1.25+ - Primary language, specializing in pure Go implementations (no CGO)
• PostgreSQL, MySQL 8.0+, SQLite - Database design and optimization
• REST APIs, WebSocket, SSE - Real-time communications

Frontend

• Angular (6-19+) - Modern component architecture
• TypeScript - Type-safe development
• NX - Monorepo tooling and architecture
• RxJS - Reactive programming patterns

Mobile

• Android - Native mobile applications

Quality Standards

• 70-95% code coverage across projects
• Zero linter issues policy
• Production-ready mindset from day one
• Comprehensive documentation

Many of the projects you see here originated from corporate development work. We've extracted reusable components, refined them into standalone libraries, and open-sourced them to benefit the community. That's why you might notice a significant number of projects launched in recent months - it's not a rush, but rather a systematic process of identifying valuable building blocks and making them publicly available.

More libraries are being extracted and will be released as open source in the coming months. This is part of our commitment to giving back to the community and fostering innovation.

We maintain and support such open source initiatives under GRPMSoft (General Resources & Power Management) - a hub for system software and infrastructure tools.

Native WYSIWYG/Rich Text editor for Angular 6-19+

A lightweight, highly customizable text editor component without external dependencies. Features include visual/HTML modes, image uploads, theming with CSS variables, and full Angular Forms integration. Launched: May 2018

Tech: Angular, TypeScript, SCSS | Status: Active development

High-performance Terminal UI framework for Go

A next-generation TUI framework featuring Domain-Driven Design architecture, perfect Unicode/emoji support, and Elm-inspired patterns. Achieves 10x performance improvement (29,000 FPS) with differential rendering. Launched: October 2025

Highlights:

• 91.8% test coverage
• 10x faster than alternatives
• Zero external dependencies
• Cross-platform mouse & clipboard support

Tech: Go 1.25+, DDD architecture | Status: Production-ready

Pure Go implementation of HDF5 file format

Modern HDF5 library with no CGO dependencies. Passes 98.2% of official HDF5 test suite (380/387 files). Supports AI/ML datatypes (FP8, bfloat16) and achieves 10-250x faster hyperslab selections. Launched: November 2025

Highlights:

• Zero C dependencies
• Full HDF5 2.0.0 compatibility
• Security hardening (4 CVEs fixed)
• 86.1% code coverage

Tech: Pure Go 1.25+ | Status: Production-ready

Production-ready race detector in pure Go without CGO

Solves the 10-year limitation where Go's race detector required C++ ThreadSanitizer. Implements FastTrack algorithm (PLDI 2009) with 260x memory optimization in common cases. Launched: November 2025

Highlights:

• Works with CGO_ENABLED=0 on all platforms
• Drop-in replacement for go build -race
• 70+ tests, 45-92% coverage
• Battle-tested in production

Tech: Pure Go, AST instrumentation | Status: Active development

Type-safe database query builder for Go with zero production dependencies

Zero production dependencies, type-safe reflection-based struct scanning. High performance with LRU statement cache and batch operations (3.3x faster). Full ACID transaction support with all isolation levels. Launched: November 2025

Features:

• PostgreSQL, MySQL 8.0+, SQLite 3.25+ support
• Expression API for type-safe WHERE clauses
• Multiple JOIN types, CTEs, subqueries
• 326+ tests, 93.3% coverage

Tech: Pure Go 1.25+ | Status: Active development

Next-generation HTTP router for Go with type-safe handlers and OpenAPI generation

Type-safe generic handlers using Box[Req, Res] pattern. High-performance routing (~10M req/s throughput, 256-326 ns/op). Automatic validation with 100+ supported tags. Launched: November 2025

Features:

• Built-in OpenAPI 3.1 spec generation
• RFC 9457 Problem Details for errors
• 8 middleware: Logger, Recovery, CORS, JWT, Rate limiting, Circuit breaker
• Zero external dependencies for core routing

Tech: Go 1.25+ | Status: Production-ready | Coverage: 93.1%

Type-safe reactive state management inspired by Angular Signals

Angular Signals pattern for Go with zero-allocation hot paths (0.51 ns/op Signal.Get). Thread-safe operations with fine-grained reactivity and glitch-free execution. Launched: November 2024

Features:

• Pure Go with zero external dependencies
• Lazy evaluation for computed values
• Effect batching for multiple updates
• Angular API compatibility

Tech: Go 1.25+ | Status: Stable | Coverage: 67.9%

Production-ready publish-subscribe library with guaranteed delivery and Dead Letter Queue

Reliable message delivery with exponential backoff retry logic and automatic Dead Letter Queue handling. Battle-tested in FreiCON Railway Management System. Supports deployment as library or standalone microservice. Launched: November 2025

Features:

• Guaranteed delivery with progressive retry delays (30s → 30m max)
• Dead Letter Queue with failure tracking and statistics
• Domain-Driven Design with Repository Pattern
• Multi-database support (MySQL, PostgreSQL, SQLite via Relica)
• Standalone REST API microservice or embeddable library
• Cloud-native with Docker support and health checks

Tech: Go 1.25+ | Status: v0.1.0 | Coverage: 95.9%

RFC-compliant Server-Sent Events and WebSocket for Go

Production-ready SSE (RFC text/event-stream) and WebSocket (RFC 6455) implementations. High performance with <100μs latency and zero external dependencies. Launched: November 2025

Features:

• Named event types, event IDs for reconnection
• Text/binary messages, control frames
• Broadcasting hub for multi-client
• 314 tests, 84.3% coverage

Tech: Go 1.25+ | Status: Production-ready

Multi-engine regex up to 263x faster than stdlib with SIMD optimizations

Multi-engine architecture (DFA/NFA/Pike VM) with AVX2/SSSE3 SIMD assembly. Automatic prefilter coordination (memchr, memmem, teddy, Aho-Corasick). Zero allocations in hot paths via object pooling. Launched: November 2025

Performance:

• Up to 263x faster than stdlib (case-insensitive patterns)
• 15-25x faster email extraction
• 30-50x faster log parsing with prefilters
• SIMD memchr: 12.3x faster on 64KB input

Tech: Go 1.25+, AMD64 SIMD | Status: v0.2.0 Experimental | Target: >70% coverage

Production-ready deep learning framework for Go with zero Python dependencies

Single-binary deployment ML framework. MNIST accuracy: 97.44% (MLP), 98.18% (CNN). Auto-differentiation via decorator pattern. Multiple backends (CPU with SIMD, GPU planned). Launched: November 2025

Features:

• Type safety with generics-powered API
• Neural Network modules: Linear, ReLU, Sigmoid, Tanh
• Optimizers: SGD with momentum, Adam
• WebAssembly support for browser inference

Tech: Go 1.25+ | Status: v0.1.1 Released | Coverage: 83.8%

Modern cross-platform shell written in Go with beautiful TUI

4 UI modes (Classic, Warp, Compact, Chat) with native POSIX script execution via mvdan.cc/sh. Git integration with branch display and dirty status indicators. Launched: October 2025

Features:

• Persistent command history with smart deduplication
• Real-time syntax highlighting
• Tab completion and multi-line input
• Viewport scrolling and visual help

Tech: Go 1.25+ | Status: Beta | Target: Q1 2026 stable

Zero-CGO WebGPU bindings — GPU-accelerated graphics and compute in pure Go

Pure Go FFI-based WebGPU bindings via wgpu-native. Cross-platform GPU access without CGO. Comprehensive API covering graphics and compute pipelines. Launched: November 2025

Features:

• Complete WebGPU API: Instance, Adapter, Device, Buffers, Textures, Samplers
• Render & Compute Pipelines with shader support
• Advanced rendering: Depth buffers, MRT, instanced/indirect rendering
• RenderBundle for pre-recorded commands
• GPU timestamp queries and error scopes
• 12 working examples (triangles, textures, 3D, compute shaders)

Tech: Go 1.25+, wgpu-native v24.0.3.1 | Status: v0.1.0 Beta | Platform: Windows/Linux/macOS

Pure Go library for reading and writing MATLAB .mat files (v5-v7.3+)

Read/write MATLAB files without CGO dependencies. Supports all numeric types, complex numbers, multi-dimensional arrays, and character arrays. Round-trip verified. Launched: June 2025

Features:

• MATLAB v5-v7.2 (binary format)
• MATLAB v7.3+ (HDF5-based format)
• Pure Go HDF5 implementation
• Cross-platform compatibility

Tech: Pure Go 1.25+ | Status: v0.2.0 Production-ready

High-performance Unicode width calculation library - 3.9-46x faster than go-runewidth

Zero memory allocations (0 B/op, 0 allocs/op) with Unicode 16.0 support. Tiered lookup strategy achieving O(1) performance for 90-95% of typical use cases. Launched: October 2025

Performance:

• 15-46x faster for ASCII
• 4-14x faster for CJK
• 6-8x faster for mixed/emoji

Tech: Go 1.25+ | Status: v0.1.0 Stable | Coverage: 87.1%

Building a production-ready ML framework for Go with zero Python dependencies.

• born - ML framework achieving 97-98% MNIST accuracy
  • Single-binary deployment, WebAssembly support
  • Auto-differentiation, multiple backends (CPU/GPU planned)
  • v0.1.1 | 132 tests, 83.8% coverage

High-performance implementations of causal discovery and statistical learning algorithms.

• causalgo - SURD algorithm for causal discovery

High-performance SURD (Synergy-Unique-Redundancy Decomposition) algorithm validated against Nature Communications 2024 research. Implements information-theoretic causal discovery with LASSO-based VarSelect for identifying causal ordering in nonlinear systems.

Highlights:

• 97.2% test coverage for SURD core
• 100% match with Python reference on turbulence data
• MATLAB file format support (v5 and v7.3 HDF5)
• Visualization exports (PNG/SVG/PDF)
• CLI tool for graph generation

Tech: Pure Go 1.25+, Gonum | Status: v0.3.0 | Coverage: 97.2%

• lasso - Parallel LASSO regression
  • Coordinate descent with goroutines
  • ~25ms on 10K samples with 100 features
  • v0.1.0 | Production-ready

Type-safe, high-performance libraries for modern Go applications.

• relica - Type-safe database query builder

  • Zero production dependencies, 3.3x faster batch operations
  • PostgreSQL, MySQL, SQLite support
  • v0.4.1+ | 326+ tests, 93.3% coverage

• fursy - Next-gen HTTP router

  • Type-safe handlers, ~10M req/s throughput
  • Built-in OpenAPI 3.1 generation, RFC 9457 errors
  • v0.1.0 | 93.1% coverage

• stream - Real-time communications

  • RFC-compliant SSE & WebSocket, <100μs latency
  • Zero external dependencies
  • v0.1.0 | 314 tests, 84.3% coverage

• signals - Reactive state management

  • Angular Signals pattern, zero-allocation hot paths
  • 0.51 ns/op Signal.Get
  • v0.1.0 | 51 tests, 67.9% coverage

• pubsub - Publish-Subscribe messaging

  • Guaranteed delivery with exponential backoff retry
  • Dead Letter Queue for failed messages
  • Battle-tested in FreiCON Railway Management System
  • v0.1.0 | 95.9% coverage, Active development

• coregex - See Featured Projects above

Pure Go WebGPU ecosystem for graphics and compute workloads.

• webgpu - See Featured Projects above

• goffi - Pure Go FFI library

  • Call C libraries without CGO
  • 88-114ns FFI call overhead
  • Platform-specific assembly optimizations
  • v0.1.0 | 87.1% coverage

GPU power, Go simplicity — zero CGO graphics ecosystem.

• gogpu - Core graphics framework

  • Zero CGO, no C compiler required
  • WebGPU backend, cross-platform (Windows/Linux/macOS)
  • Simple API inspired by raylib, Processing, Ebitengine
  • WIP | Project structure ready

• wgpu - Pure Go WebGPU implementation

  • No wgpu-native, no Rust dependencies
  • Direct GPU access: Vulkan, Metal, DirectX 12, OpenGL
  • WASM-compatible for browser deployment
  • Early | Long-term vision

• gg - GPU-accelerated 2D graphics

  • Simple API like fogleman/gg but GPU-powered
  • Processing-style creative coding interface
  • Shapes, curves, text, images, gradients
  • Planned | After gogpu v1.0

• naga - Pure Go shader compiler

  • Port of Rust naga to pure Go
  • WGSL → SPIR-V, GLSL, HLSL, MSL
  • Runtime shader generation, WASM support
  • Coming Soon | Active development

• ui - Pure Go GUI toolkit

  • Modern widgets: buttons, inputs, tables, modals
  • Flex/Grid/Stack layouts, themeable (dark/light)
  • GPU-accelerated 60fps, cross-platform
  • Planned | After gogpu v0.2.0

Building Go's scientific computing ecosystem with pure Go implementations.

• hdf5 - See Featured Projects above

• matlab - MATLAB file format library

  • Read/write .mat files (v5-v7.3+)
  • Pure Go, no CGO
  • v0.2.0-beta | Active development

• phoenix - See Featured Projects above

• gosh - Modern cross-platform shell

  • 4 UI modes: Classic, Warp, Compact, Chat
  • Native POSIX script execution, Git integration
  • v0.1.0-beta.7 | Targeting stable Q1 2026

• grpm - 🔒 Private (launching Q1 2026)

  • General Resource & Package Manager
  • Universal package management for Go ecosystem

• iris - 🔒 Private (launching Q2 2026)
  • IrisMX modern mail exchange server
  • Modern SMTP/IMAP/POP3 implementation

• uniwidth - Unicode width calculation
  • 3.9-46x faster than go-runewidth
  • Zero allocations
  • v0.1.0-beta | 87.1% coverage

• ngx-dadata - DaData address autocomplete for Angular

• iso6346 - Container number validation (ISO 6346)
• luhn - Luhn algorithm (credit card validation)
• esr - Check-digit mod11 algorithm

• url-translit - Cyrillic to Latin URL transliteration
• prerender - Headless Chrome prerendering server

• webgpu v0.1.0 - Public release
• uniwidth v0.1.0 - Stable release
• born v0.2.0 - RNN/LSTM architectures

• gogpu v0.1.0 - Basic rendering framework
• grpm v0.1.0-beta - Public beta
• gosh v1.0.0 - Stable shell release
• signals v0.2.0 - Advanced features
• phoenix v0.2.0 - Signals integration

• gg v0.1.0 - GPU-accelerated 2D graphics
• born v0.3.0 - Attention mechanisms
• iris v0.1.0-beta - Mail server public beta
• relica v1.0.0 - Stable API
• fursy v1.0.0 - Stable API

• naga v0.1.0 - Pure Go shader compiler
• born v0.5.0 - GPU backends (CUDA, Vulkan, Metal)
• hdf5 v1.0.0 - Full HDF5 compliance
• racedetector v1.0.0 - Stable release
• grpm v0.2.0 - Production ready

• ui v0.1.0 - Pure Go GUI toolkit
• gogpu v1.0.0 - Stable graphics framework
• born v1.0.0 - Production ML framework

I'm available for consulting and contract work in my areas of expertise:

• Go Backend Development - Microservices, APIs, system programming
• Angular/TypeScript Frontend - SPAs, component libraries, reactive architectures
• Performance Optimization - Profiling, bottleneck analysis, zero-allocation strategies
• Code Review & Architecture Audit - Best practices, scalability assessment
• Open Source Contributions - Happy to contribute to projects serving the public good

• Pure Go implementations (no CGO)
• Type-safe APIs with Go generics
• High-performance systems (TUI, databases, networking)
• Scientific computing in Go
• Angular component development
• Database query optimization
• WebAssembly and cross-compilation

• Part-time contracts
• Freelance projects
• Technical consulting
• Code review services
• Open source collaboration

I believe in:

• Quality over quantity - Every project should be production-ready
• Zero compromises - Maintain high code coverage and zero linter issues
• Pure Go - Avoid CGO dependencies for better cross-compilation
• Type safety - Leverage Go 1.25+ generics for compile-time guarantees
• Community first - Clear documentation, examples, and responsive maintenance
• Enterprise to Open Source - Extracting battle-tested components from production systems and sharing them with the community

Many projects here have been refined in real-world corporate environments before being released as open source. This ensures they're not just theoretical solutions, but proven tools that solve actual problems. We continue this practice through GRPMSoft and other organizations, systematically identifying and open-sourcing valuable components.

• GitHub: @kolkov
• Organizations: born-ml | CausalGo | coregx | go-webgpu | scigolib | phoenix-tui | grpmsoft | unilibs

Open to: Consulting opportunities, open source collaboration, technical partnerships

• Building a comprehensive ML ecosystem for Go (born-ml)
• Advancing scientific computing libraries in pure Go (scigolib)
• Creating high-performance web infrastructure (coregx)
• Developing modern system tools (grpmsoft)
• Expanding GPU computing capabilities in Go (go-webgpu)