I build Telegram/VK bots, REST APIs, and automation for small businesses: booking, leads, catalogs, Sheets.
Python is the foundation. Backend is the core. Bots and AI are how the process reaches the client.
| Name | Oleg |
| Role | Python Backend Developer |
| Specialization | Bots + AI automation for SMB |
| Location | Russia |
| Open to | Freelance · Remote roles |
| Languages | Russian (Native) · English (Intermediate / Technical Reading) |
- Business bots (booking, leads, shop)
- Python backend systems
- REST APIs
- AI assistants and RAG
- Telegram and VK bots
- Google Sheets / Drive / Calendar automation
- Application intake (nifontov.tech)
| Area | Technologies |
|---|---|
| Languages | Python, SQL, Bash |
| Backend | FastAPI, Flask, aiogram 3, SQLAlchemy, Pydantic, asyncio |
| Databases | SQLite, PostgreSQL |
| AI | OpenAI API, LangChain, LangGraph, Pinecone, GigaChat API |
| APIs & integrations | REST APIs, Telegram Bot API, VK API (vkbottle), Google Sheets / Drive / Calendar |
| Tools | Git, GitHub, Linux, VS Code, PyCharm |
Desktop mini-CRM with a FastAPI REST backend, SQLite storage and a Tkinter UI. Covers clients, deals, orders and tasks, with export to Google Sheets / Drive and styled KPI spreadsheets.
Stack: Python · FastAPI · Pydantic · SQLite · Google Sheets API · Google Drive API
RAG Telegram bot with Pinecone vector search, URL ingestion into embeddings, LangChain agent tools, per-user memory and an external REST tool.
Stack: Python · LangChain · Pinecone · OpenAI-compatible API · Telegram Bot API
Telegram shop agent split into an aiogram 3 bot and a FastMCP server. Catalog, cart and checkout live in SQLite; the model routes tool calls over MCP, with extra tools for weather, crypto and other public APIs.
Stack: Python · aiogram 3 · FastMCP · SQLite · OpenAI-compatible API
Telegram AI assistant for a barbershop: dialogue over a service knowledge base, appointment booking with slot checks, admin notifications and optional Google Calendar sync. Function calling drives booking actions.
Stack: Python · python-telegram-bot · GigaChat · SQLite · Google Calendar API (optional)
VK community assistant: GPT chat, 4-step consultation booking, FAQ and quiz on vkbottle.
Stack: Python · vkbottle · OpenAI-compatible API · JSON storage
Application form — nifontov.tech
Lead intake: React form, FastAPI, PostgreSQL. Qualifies the brief. Not a portfolio gallery.
Stack: FastAPI · PostgreSQL · React · Docker
Pinned repository cards
| Project | Focus |
|---|---|
| langchain-content-pipeline | 4-step LangChain: analysis → tools → generate → review |
| flask-go-users-api | Users REST API in Flask (and a parallel Go service), SQLite, OpenAPI 3.1 |
| langchain-telegram-tool-agent | LangChain tool-calling agent: search, weather, FX/crypto, files, HTTP, QR |
| dori-nemo-memory-bots | Two aiogram 3 bots comparing in-memory context vs PostgreSQL facts (SQLAlchemy 2) |
| telegram-summary-bot | Collects Telegram messages into SQLite and builds GigaChat summaries; Flask viewer |
| haystack-telegram-agent-v2 | Haystack agent with Pinecone memory, document RAG and OpenAI tools |
| langchain-breed-agent | LangChain / LangGraph agent with structured output and conversation memory |
| vk-weather-bot | VK weather bot: FSM, forecast, geolocation and air quality via OpenWeatherMap |
| dual-llm-text-agent | CLI chat agent with dialogue memory: Claude thinking mode and OpenAI models |
| pdf-generator-from-data | PDF generation from CSV, JSON and Google Sheets using HTML templates |
| Direction | Where it shows up |
|---|---|
| AI assistants | Knowledge-base dialogue, booking and FAQ in Telegram / VK |
| AI agents | Tool-calling agents (LangChain, Haystack, MCP) that call APIs and local tools |
| RAG | Pinecone semantic search, URL/document ingestion, hybrid group memory |
| LLM pipelines | Multi-step LangChain chains with Pydantic structured output |
| Function calling | Booking slots, catalog/cart actions, weather and other external tools |
| API automation | Google Sheets / Drive / Calendar, OpenWeather, public HTTP APIs |
| Business workflows | CRM records and reports, salon booking, shop checkout, chat summarization |
| Area | Practice in public repos |
|---|---|
| Python | Async bots, CLI tools, REST services |
| REST API | FastAPI CRM API; Flask users API with OpenAPI 3.1 |
| Databases | SQLite across most apps; PostgreSQL + SQLAlchemy 2 in memory-bot backends |
| External APIs | OpenAI-compatible LLMs, Google APIs, Telegram, VK, weather and market APIs |
| Channel | Details |
|---|---|
| Website | nifontovv.ru |
| Application form | nifontov.tech |
| GitHub | github.com/nifontovoleg |
| o.nifontov@yandex.ru | |
| Telegram | @olegugfv_reg59 |