Full-Stack · Automation · AI Systems · Infrastructure

I build systems that replace manual work.

Websites. Automation. Bots. AI. Infrastructure.

If a process can be programmed, connected to other systems, automated or improved with AI — I design and build the system around it: from interface and backend to the server it runs on.

See what I can automate
  • 10+years of programming
  • 5layers: UI → API → DB → AI → infra
  • 24hreply to your request — within a day
Website → API → Backend → Database → AI → Automation → Telegram → AnalyticsWebsiteAPIBackendDatabaseAIAutomationTelegramAnalytics

One system, every layer — and it runs on its own.

What I build

From idea to working system.

Not a single page or a single script — a working product: interface, backend, data, automation and the server it runs on.

Websites

Frontend and backend as one whole — from a landing page to a marketplace with an admin panel.

  • Landing pages
  • Corporate websites
  • E-commerce
  • Marketplaces
  • Admin panels
  • Dashboards
  • Web applications

Process

How I build digital products.

The same path for a landing page and for a system with AI inside. The only difference is how deep each step goes.

  1. 01UnderstandThe task, the process as it is today, constraints, the definition of done.
  2. 02ArchitectArchitecture, data, integrations, the points where automation belongs.
  3. 03DesignStructure, UX, interface — built around the real user scenario.
  4. 04BuildFrontend, backend, database, API.
  5. 05AutomateParsers, schedulers, bots, AI — everything that repeats.
  6. 06TestTests, browser checks, screenshots, edge cases.
  7. 07DeployServer, Docker, Nginx, backups, monitoring.
  8. 08ImproveLogs, metrics, feedback — and the next iteration.

The same process — AI-assisted

Some steps are carried out by AI agents under my supervision. The result is verified in a real browser — with a screenshot, not on faith.

Working process · AI agents + browser verification
  1. Prompt
  2. Research
  3. Design
  4. Code
  5. Playwright
  6. Screenshot
  7. Visual QA
  8. Fix
  9. Deploy

Automation

What can be automated?

Almost everything that repeats. Here is how a routine that eats an employee's day turns into a system that runs by itself.

Before

An employee, every day

  1. 01opens the website
  2. 02copies the information
  3. 03puts data into a spreadsheet
  4. 04checks the prices
  5. 05sends a notification
  6. 06updates the database

~2–4 hours a day · errors · delays

After

One automated system

Parser
API
Database
AI
Telegram
Dashboard

runs on schedule · 0 manual steps · logs everything

Typical tasks I automate

  • Competitor price monitoringScheduled parsing, price change detection, alert to Telegram.
  • Automatic catalog updatesSupplier feed → normalization → your database → storefront.
  • Telegram alertsOrders, errors, server events — straight to the team's chat.
  • Order processingForm → validation → CRM → notification → status tracking.
  • Content generationDescriptions, titles and images generated by AI from structured data.
  • Data collectionThousands of pages into one clean table with deduplication.
  • ReportingDaily/weekly reports assembled and delivered without a human.
  • Browser automationLogins, forms, clicks — a headless browser instead of a person.
  • AI processingClassification, extraction, summaries over incoming data.
  • Scheduled synchronizationTwo systems kept in sync by cron, with retries and logs.

Have a repetitive task? Let's automate it.

Describe the task

How it's built

Systems, not scripts.

Five typical circuits that most of my projects are made of. Each can be ordered on its own or as part of a whole system.

01

Data & parsing

A 'parser' is not a one-page script. It is a pipeline: many sources, normalization, validation, a database — and consumers that need clean data.

  1. Sources
  2. Ingestion
  3. Normalization
  4. Validation
  5. Database
  6. API
  7. Website · Telegram · Dashboard
  • Catalogs of tens of thousands of items
  • Supplier feeds: JSON, CSV, XML
  • Deduplication, filtering, transformation
  • Scheduled synchronization, price monitoring
  • Aggregation from several sources into one data model
02

Browser automation

When a service has no API, it can still be operated through the browser — like a person would, but on a schedule and with the result verified.

  1. Browser
  2. Playwright
  3. Website
  4. Action
  5. Validation
  6. Screenshot
  7. Result
  • Navigation, forms, authenticated scenarios
  • Recurring operations in web interfaces
  • Data collection from JS-rendered sites
  • Screenshots and visual verification of interfaces
  • E2E tests and availability monitoring

Only where it is technically and legally appropriate. Bypassing security mechanisms is not what this is about.

03

Telegram as an interface

Not 'just a bot' but an entry point into a system: orders, alerts, admin tools and reports — where the team already is.

  1. Website
  2. Backend
  3. Order
  4. Telegram notification
  1. Monitoring
  2. Trigger
  3. Telegram alert
  1. User
  2. Bot
  3. API
  4. Database
  5. Response
  • Customer and admin bots
  • Notifications and event monitoring
  • Order intake and management
  • Support automation
  • Internal tools for the team
04

Integrations

I connect systems that were never designed to work together: CRM, store, payments, delivery, analytics, Telegram, internal databases.

  1. Service A
  2. API / Webhook
  3. Automation layer
  4. Database
  5. Service B
Ways to connect
  • API
  • Webhooks
  • Database
  • Files
  • Parser
  • Browser
  • Scheduled sync
  • CRM ↔ website ↔ Telegram
  • Payments, delivery, inventory
  • Analytics and reporting from several sources
  • External APIs and internal systems
05

Infrastructure

The product lives on a server that I set up and maintain myself: from reverse proxy to backups and GPU workloads.

  1. Internet
  2. Nginxreverse proxy · TLS · static
  3. Application
    • Backend
    • Database
    • Automation
    • AI
    • Workers
  4. Docker · Linuxservice isolation · compose
  5. ServerUbuntu · backups · monitoring
  • Ubuntu Server, Docker Compose, Nginx
  • Deployment, SSH, Git-based workflow
  • Databases, queues, background workers
  • Backups, monitoring, service management
  • GPU workloads and model serving

Lab

Don't take my word for it.

Small interactive demos. Each one is a simplified version of something I actually build. Everything is simulated in your browser and not connected to live systems.

Parser

Interactive demo

URL → HTML → structured data → JSON

  1. Fetch page
  2. Extract fields
  3. Normalize
  4. JSON
// JSON

Safe demo: parses a prepared page, never an arbitrary site.

Automation

Interactive demo

One click — the whole chain runs.

Source
Parser
Database
AI
Telegram
$ 

API

Interactive demo

A clean REST API, the way I build them.

// response

AI assistant (RAG)

Interactive demo

How an AI layer looks inside a product: question → search over your own data → answer.

  1. Question
  2. Search
  3. Qdrant
  4. Context
  5. Model
  6. Answer

Mock interface — answers are scripted, no data leaves the page.

Browser QA

Interactive demo

Build → browser → screenshot → analysis → fix. This is how a result gets verified instead of taken on faith.

  1. Build
  2. Browser
  3. Screenshot
  4. Analysis
  5. Fix
$ 

Dashboard

Interactive demo

Live analytics — the kind an owner opens each morning.

Orders452
Revenue808 628
Conversion3.4%
Uptime99.98%
Orders, last 14 days
-14d-12d-10d-8d-6d-4d-2d

Terminal

Interactive demo

A typical deploy — simulated, safe.

deploy — ssh

System

Every layer, one pair of hands.

Hover or tap a node to see what I can build there.

Stack

Tools I actually use.

Grouped by layer. Click a group to expand.

Developer

Frontend

  • HTML
  • CSS
  • JavaScript
  • TypeScript
  • React
  • Astro
  • Three.js

AI Lab

I run my own AI infrastructure.

Local models on my own GPU server — no per-token bills, no data leaving the network. The same setup can be deployed for your product.

GPU Server

Models

LLM
Image
Video
Embeddings
Agents
API
Applications
  • Local LLMOllama-served models for chat, extraction, classification and agents.
  • AI agentsTool-calling agents that read your data, call your API and act.
  • Vector databasesQdrant + embeddings: semantic search and RAG over documents and catalogs.
  • Image generationComfyUI pipelines for product visuals, variations and upscaling.
  • Video workflowsGeneration and processing pipelines for short-form video.
  • AI-assisted developmentAI agents as part of the development workflow — faster delivery, same rigor.

How AI lives inside a product

Not 'plugged in a chatbot' but a dedicated layer: a model, retrieval over your data and tools that act inside your system.

User
Application
AI API
LLM
RAG
Tools
Qdrant
API / Database
Response

AI media

Image and video generation and processing on my own GPU — ComfyUI workflows that can be embedded into a pipeline.

  • ImageProduction
    • Generation
    • Editing
    • Product visuals
    • Reference-based workflows
    • Enhancement & upscale
  • VideoExperimental
    • Generation
    • Image-to-video
    • Workflow orchestration
    • Local video models
    • Automated pipelines
  • AudioResearch
    • Speech synthesis
    • Voiceover
    • Audio processing

Honest labels: 'production' — I use it in real tasks, 'experimental' — works but not production-ready yet, 'research' — still exploring.

Tools and their role

Not logos for the sake of logos — what each tool actually does in my systems.

  • OllamaLocal LLM serving
  • QdrantVector retrieval / knowledge search
  • ComfyUINode-based AI media workflows
  • PlaywrightBrowser automation / testing
  • DockerService isolation / deployment
  • AI agentsAssisted development in the terminal
  • GitVersioning and deployment
  • UbuntuOS for servers and the GPU workstation
Experimental architecture · in development

My AI development platform

An environment where AI doesn't just write code — it sees the result.

I am building a local AI-assisted development environment: models, a knowledge base, role-based agents, browser automation and visual verification — on my own infrastructure.

  • AI models
  • Knowledge base
  • Development agents
  • Design intelligence
  • Browser automation
  • Visual verification
  • Infrastructure
User request
Orchestrator
  • Architectstructure, data, integrations
  • Researchsources, documentation
  • Designcomposition, patterns
  • Codeimplementation
  • Automationscripts, pipelines
  • Testbrowser, checks
  1. 01Build
  2. 02Browser
  3. 03Playwright
  4. 04Screenshot
  5. 05Visual analysis
  6. 06Issue detection
  7. 07Fix / iteration
  8. 08Result

Visual verification loop

Instead of 'generated the code and assumed the interface is correct' — the environment opens the result in a browser, looks at it and iterates.

  1. Build
  2. Open
  3. Screenshot
  4. Analyze
  5. Issues
  6. Fix
  7. Retest
  • Playwright
  • Screenshot analysis
  • Vision model
  • DOM inspection
  • Automated tests

Design intelligence

A dedicated layer of knowledge about interfaces, so the agent composes pages from proven patterns rather than 'whatever comes out'.

  1. Task
  2. Design retrieval
  3. Relevant components
  4. Page composition
  5. Implementation
  6. Screenshot
  7. Visual critique
  8. Improvement
  • UI libraries
  • Design systems
  • Components
  • Landing pages
  • E-commerce patterns
  • Forms
  • Navigation
  • Dashboards
  • Typography
  • Spacing & grids
  • Motion
  • 3D
  • Modern web patterns

The base is curated references, documentation, open resources and pattern metadata. Not copies of commercial UI libraries.

What already works, and what doesn't yet

Working
  • AI-assisted development with agents
  • Browser automation and screenshots (Playwright)
  • Local models, Qdrant, ComfyUI on my own GPU
In development
  • Orchestrator and project knowledge base
  • Design layer: pattern and component retrieval
  • Automatic screenshot → analysis → fix loop
Experimental
  • Specialized agents as separate roles
  • Vision model as a visual critic
  • Fully autonomous iterations without a human

Code

This is what my code looks like.

Realistic fragments — API integration, automation, data processing. No secrets, no hello-worlds.

api/webhooks/order.phpOrder webhook → validation → DB → Telegram
<?php
declare(strict_types=1);

// Webhook нового заказа: подпись → валидация → БД → уведомление.
// Секреты берутся из окружения, в коде их нет.

$secret  = getenv('ORDER_WEBHOOK_SECRET') ?: '';
$payload = file_get_contents('php://input');
$given   = $_SERVER['HTTP_X_SIGNATURE'] ?? '';

if (!hash_equals(hash_hmac('sha256', $payload, $secret), $given)) {
    http_response_code(401);
    exit;
}

$order = json_decode($payload, true, 8, JSON_THROW_ON_ERROR);

$rules = [
    'id'     => fn($v) => is_int($v) && $v > 0,
    'phone'  => fn($v) => preg_match('/^\+?\d{10,15}$/', (string) $v),
    'items'  => fn($v) => is_array($v) && count($v) > 0,
    'total'  => fn($v) => is_numeric($v) && $v >= 0,
];
foreach ($rules as $field => $ok) {
    if (!isset($order[$field]) || !$ok($order[$field])) {
        http_response_code(422);
        echo json_encode(['error' => "invalid {$field}"]);
        exit;
    }
}

$db = Db::connection();
$db->beginTransaction();
try {
    $db->prepare(
        'INSERT INTO orders (id, phone, total, status, created_at)
         VALUES (:id, :phone, :total, "new", NOW())
         ON DUPLICATE KEY UPDATE total = VALUES(total)'
    )->execute([
        'id' => $order['id'], 'phone' => $order['phone'], 'total' => $order['total'],
    ]);

    $stmt = $db->prepare('INSERT INTO order_items (order_id, sku, qty) VALUES (?, ?, ?)');
    foreach ($order['items'] as $item) {
        $stmt->execute([$order['id'], $item['sku'], (int) $item['qty']]);
    }
    $db->commit();
} catch (Throwable $e) {
    $db->rollBack();
    error_log('order webhook: ' . $e->getMessage());
    http_response_code(500);
    exit;
}

// Уведомление — асинхронно, чтобы не держать вебхук открытым.
Queue::push('telegram.notify', [
    'chat' => 'orders',
    'text' => sprintf("🛒 Заказ #%d — %s ₴\n%d позиций",
        $order['id'], number_format($order['total'], 0, '.', ' '), count($order['items'])),
]);

http_response_code(202);
echo json_encode(['ok' => true]);

Experience

10+ years

Programming & web development.

Each layer was added because the previous one wasn't enough to ship a whole product.

  1. Web development

    Sites, layouts, CMS, first backends in PHP.

  2. Backend

    APIs, databases, payments, auth — logic behind the interface.

  3. Automation

    Parsers, bots, schedulers — removing manual work from processes.

  4. Infrastructure

    Own servers: Linux, Docker, Nginx, backups, monitoring.

  5. AI systems

    Local models, agents, RAG — AI as a layer inside products.

About

Hi. ggcode13.

I've been programming for more than 10 years. I work across the full product stack — interface, backend, databases, automation, servers and AI.

I enjoy problems where software can remove repetitive work or turn a complicated process into a simple system.

Don't automate because it's trendy. Automate because humans shouldn't waste time doing what software can do reliably.
  • 01BuildA working product first — interface, backend, data.
  • 02AutomateThen remove every step a human repeats.
  • 03ScaleThen make it run on its own server, reliably.

Contact

Have a problem that software can solve?

Tell me what you're trying to automate, build or improve. You don't need to know which stack you need — just describe the task.