Switch to English?
Yes
Переключитись на українську?
Так
Переключиться на русскую?
Да
Przełączyć się na polską?
Tak
  1. Freelancers

Volodimir Maryuha

Offer Volodimir work on your next project.

Ukraine Rovno, Ukraine
26 days 17 hours back
Available for hire available for hire
age 34 years
on the service 1 year
Freelancer Volodimir M. — Ukraine, Rovno. Core skill — AI Content Creation, Cybersecurity & Data Protection

Rating

Successful projects
No data
Average rating
No data
Rating
232
AI Content Creation 1
110 place out of 1551
Cybersecurity & Data Protection
30 place out of 197

Language proficiency level

Українська Українська: fluent
English English: pre-intermediate

Skills and abilities

Design & art

Services

Photo, Audio & Video


Promotion

Mobile development

Translation


Writing

Portfolio


  • Gomin News — Telegram bot for automatic aggregation and AI processing

    Bot Development
    Developed a production-ready Telegram bot for automatic collection, AI processing, and publication of news in a Telegram channel.

    #telegram_bot #python #AI #automation #news

    … WHAT THE BOT DOES:
    — Collects news from sources in the background (background worker)
    — Processes content through AI: reframing, summarization, formatting
    — Stores processed entries in SQLite without duplicates
    — Automatically publishes ready posts in the channel(s)
    — Supports multi-channel logic

    TECHNOLOGIES:
    — Python (modular architecture: handlers / services / utils)
    — SQLite for state storage
    — Docker + docker-compose for deployment
    — Telegram Bot API
    — AI integration for text processing

    RESULT:
    A fully autonomous service that operates 24/7 on a server and publishes processed news daily without human involvement.

  • Artemis — personal AI assistant for Telegram

    Bot Development
    Developed a personal Telegram bot in Python with a multimodal AI router: Gemini 3 Flash as the main brain and Groq Compound as a fallback with automatic switching during failures.

    WHAT THE BOT CAN DO:

    … - Unified Brain: one AI call analyzes intent, mood, facts, and action type simultaneously
    - Image generation from text description (gemini-2.5-flash-image)
    - Voice messages: TTS voiceover and voice recognition via Whisper
    - Photo analysis sent in chat (vision)
    - Long-term memory: stores facts, preferences, and user events in SQLite
    - Journal with timestamps
    - Game progress tracking (Roblox, CS2)
    - Smart reminders: AI extracts time from natural language
    - Agent tasks: the bot independently searches for information and prepares a response for the required time
    - Personalized news by subscription on topics, delivered on schedule
    - APScheduler: one-time and cyclical tasks in real-time

    #telegram #python #ai #gemini #groq #bot

  • Smakota — Telegram bot + Mini App for food delivery automation

    Bot Development
    Developed a full-fledged ecosystem for restaurant business automation in Python — from customer ordering to delivery confirmation by the courier.

    WHAT THE SYSTEM CAN DO:

    … — Telegram Mini App (WebApp) — a full menu website directly in Telegram with a cart, dish photos, categories, and order placement
    — Customer orders through WebApp or via the bot, auto-filling data from previous orders
    — 5 roles: customer, dispatcher (inputs phone orders for customers into the bot), hall waiter, courier, admin/chef
    — Hall waiter opens checks for tables
    — Courier sees the route with a Google Maps button, manages shifts (on/off)
    — Batch routing: admin forms a route from multiple orders and sends it to the courier with one click
    — Real-time courier monitoring
    — Menu updates from Google Sheets with one command /updatemenu (asynchronously, thread-safe)
    — Mailings to all customers with templates
    — Purchase list for the kitchen
    — Daily revenue for the chef

    TECHNOLOGIES:
    Python, pyTelegramBotAPI, SQLite, Google Sheets API, HTML/CSS/JS (Telegram WebApp), GitHub Pages

    #telegram #bot #python #webapp #delivery #restaurant #automation

  • SynaptoClaw — cognitive memory plugin for AI agents

    AI & Machine Learning
    SynaptoClaw is an open cognitive memory plugin for AI agents on the OpenClaw platform, written in TypeScript.

    WHAT THE SYSTEM DOES:

    … Unlike standard RAG solutions that use simple vector similarity, SynaptoClaw replicates key neurological functions of the human brain.

    WORKING MEMORY (BUFFER): Facts initially enter the buffer and are moved to long-term storage (LanceDB) only when exceeding importance thresholds or upon repetition - analogous to the Hippocampus.

    ASSOCIATIVE THINKING (AMHR): The knowledge graph allows finding related facts even with low vector similarity.

    7-CHANNEL MEMORY SCORING: Vector similarity, importance, relevance, temporal context, reinforcement, graph connections, emotional tone.

    REFLECTION: Generation of a psychological profile of the user based on all stored memories.

    MEMORY CONSOLIDATION: Merging duplicates via CLI.

    CONVERSATION STACK: 17x token compression while preserving context.

    TECHNOLOGIES: TypeScript, LanceDB, Google Gemini API, OpenAI API, Knowledge Graph, JSONL Observability Tracer, Vitest.

    #AI #TypeScript #MemoryPlugin #OpenClaw #KnowledgeGraph #MachineLearning #NodeJS

  • Thought writing — from voice to structured thinking

    App Development for Android
    Thought writing — from voice to structured thinking — example of freelancer's portfolio work in category App Development for Android
    Brief about the application

    - Package name (`pubspec.yaml`): `dumkopys`
    - Platforms: Android, iOS, Web, Windows, macOS, Linux
    … - State: Riverpod (Notifier + manual providers)
    - Recording: `record` + `audioplayers`, local audio storage
    - Transcription: Google Gemini via backend proxy (mostly) or direct API (fallback)
    - Authentication: Firebase Auth (Google Sign-In), secure initialization
    - Data: Firestore for history (if logged in), `SharedPreferences` for local settings/history
    - Remote Config: retrieves `gemini_api_key` and `gemini_endpoint` with priorities and fallback

    Key functionality and UX.

    - Audio recording with live volume visualization.
    - Transcription via Gemini (raw and strict JSON with header).
    - History of transcriptions: renaming, restoring, deleting.
    - Haptics: tactile feedback during states.
    - "Quick copy" and sharing through system mechanisms.
    - Themes: light/dark/system, custom fonts.
    - Web, Android, iOS, desktop targets (web audio limitations depend on the browser).

  • Automation of daily plan setting for the sales team

    Enterprise Resource Planning (ERP)
    Problem: Sales representatives were spending time and mental energy daily calculating their individual sales plans to meet the monthly quota. This led to calculation errors and distracted them from the main task – sales.
    My solution: I created an automated system in MS Excel. Instead of forcing people to calculate, the system analyzed current sales and provided each employee with a ready, clear target for the day. I simply sent them the final figures.
    Result:
    Completely removed mental load and routine calculations from the team.
    … Freed up 15-20 minutes daily for each employee, which they could spend preparing for visits.
    Eliminated 100% of planning errors, making the process completely transparent.

Activity

  Latest proposals 2
Telegram bot for collecting applications from advertising (simple scenario, without CRM)
22 USD
A massage salon needs a bot in Telegram.
22 USD