Constantine
RU EN 中文
konstantin@dev:~$ whoami fullstack developer · Python / JS · 6+ years

I speak the language
of business and turn it
into a working product

Web apps, online stores, bots and automations — from the first conversation and spec all the way to production. Six years of development plus several years of project management: between “what the business needs” and “how to build it” I need no translator.

experience
6+ years in commercial development
integrations
20+ external APIs: CRM, payments, social, exchanges
english
C1 — experience on international projects
status
available for projects

~/services $ ls --all

Full-cycle web development — for your goal

From a simple business-card site to a high-load web app. Grouped by area below; in practice I take on anything web, data and automation related.

01

Sites & landing pages

Business-card sites, landing and promo pages. Fast, responsive, SEO-ready with tasteful motion.

  • Next.js
  • Tailwind
02

Web apps & SaaS

Customer portals, dashboards, analytics services and portals with auth and role management.

  • Django
  • FastAPI
  • React
03

Online stores

Catalog, cart, checkout and payments, integrations with CRM, inventory and delivery.

  • Django
  • API
  • CMS
04

Payments

Local and international cards, crypto payments. Subscriptions, balances, webhooks, idempotency.

  • Stripe
  • Plisio
  • YooKassa
05

Bots & automation

Telegram bots, routine automation, data parsing and scraping with anti-bot handling.

  • aiogram
  • Celery
  • Playwright
06

API integrations

Connecting services: CRMs, payment providers, social platforms, exchanges and any external API — reliably, with retries and rate-limit handling.

  • REST
  • WebSocket
  • webhooks
07

AI / LLM solutions

Assistants, RAG over your knowledge base, content generation, data classification and analysis.

  • Claude
  • OpenAI
  • RAG
08

Analytics & data

Data collection and processing, real-time dashboards, CSV and PDF reports.

  • PostgreSQL
  • Redis
  • Elastic

~/about $ cat README.md

An engineer with a product mindset

Python developer focused on AI and analytics. I dig into a problem until I truly get it, build a working solution and polish it to a clean result.

Before development I spent several years running projects and leading a team: gathering requirements, writing specs, coordinating development and design, accepting the result. So I work as comfortably with people as with code — I understand why a feature is needed, not just how to build it.

I do my best work on independent projects: my own area of responsibility, clear deadlines, varied tasks. I organize the process myself, without micromanagement, and own the outcome — not the hours logged.

git diff — what I add to a project

A bridge between business and code
I translate business goals into technical solutions and align both sides.
Analytical thinking
I get to the root of things and notice what others usually miss.
Outcome-driven
Practicality over theory for its own sake — the tangible result matters.
Self-reliant
I run a project end to end, keep deadlines and need no micromanagement.
Non-obvious solutions
Where “the usual way” is awkward, I look for a smarter path.
Fluent English (C1)
Experience working with international teams and clients.

~/stack $ cat requirements.txt

The stack I work with

Battle-tested in production on SaaS platforms, trading terminals and analytics services.

backend
  • Python
  • Django
  • FastAPI
  • Channels
  • Celery
  • REST
  • WebSocket
frontend
  • JavaScript
  • React
  • Next.js
  • Tailwind
  • HTMX
  • Alpine.js
ai / llm
  • Claude
  • OpenAI
  • Gemini
  • RAG
  • Prompt Engineering
  • Ollama
data / real-time
  • PostgreSQL
  • Redis
  • Elasticsearch
  • WebSocket streams
  • pandas
payments
  • Stripe
  • Plisio
  • webhooks
  • idempotency
devops
  • Docker Compose
  • Linux
  • gunicorn / daphne
  • Git
  • Sentry
  • pytest

~/work $ ls -l --sort=impact

Selected projects

Production services I built and maintained.

twtdata.com · SaaS · Twitter / X analytics live

A Twitter/X analytics platform: data collection and indexing, search, reports, payments and multilingual content.

  • Twitter/X API integration with token rotation, circuit breaker and rate-limit handling
  • Search and aggregations in Elasticsearch, streaming CSV and PDF reports
  • Card (Stripe) and crypto (Plisio) payments with idempotent webhooks
  • LLM-generated SEO content across 9 languages
  • Django
  • Celery
  • Elasticsearch
  • Stripe
  • LLM
Open site ↗
tkmetrics.com · SaaS · TikTok analytics live

A TikTok analytics service: multi-source data collection with fallback, engagement analytics and reports.

  • Data collection with fallback: Apify, a public API and a self-hosted headless scraper with anti-bot handling
  • Engagement analytics, tiers and time-series snapshots on Django ORM + Redis
  • Stripe payments with a prepaid balance and idempotent webhooks
  • Async CSV exports and periodic tasks on Celery
  • Django
  • Playwright
  • Redis
  • Stripe
  • Celery
Open site ↗
Automated trading terminal · trading + AI under NDA

Automated crypto-futures trading: real-time market data, LLM analysis, risk management and trade execution.

  • Bybit WebSocket streams with auto-reconnect; subscriptions via Redis pub/sub, REST fallback
  • Multi-tier LLM decision pipeline (filter → decision) via OpenRouter and Claude
  • Risk-management engine: 25+ validation gates, kill switch, manual confirmation
  • Real-time dashboard: Django Channels, lightweight-charts, htmx/Alpine/Tailwind
Terminal main screen: watchlist, candlestick chart, Tier 2 decisions, open positions, news feed and balance Positions table: leverage, size, SL and TP, trailing, PnL and the model behind each decision AI decision log (Tier 2): action, confidence, model and a written rationale for every decision LLM call log: purpose, model, tokens, cost, latency and request status LLM cost statistics: hourly, daily, weekly and monthly slices, broken down by model and purpose LLM model registry: providers, labels and switching the active models
01 / 06 Terminal main screen: watchlist, candlestick chart, Tier 2 decisions, open positions, news feed and balance
  • Django
  • Channels
  • Bybit API
  • LLM
  • pandas

~/process $ git log --reverse

A transparent process from idea to launch

  1. 01

    Discovery

    We dig into the task and goals. I ask the right questions to understand what the business actually needs.

  2. 02

    Spec & estimate

    I lock scope, timeline and cost. You know upfront what you get and when.

  3. 03

    Development

    I build in iterations and show progress, keeping you in the loop in plain language.

  4. 04

    Launch & support

    I ship to production, test and hand over — and support or extend it further if needed.

~/formats $ cat *.md

Three ways to get started

mode 01

Turnkey project

From idea to launch: a site, store, web app, bot or automation — fully under my responsibility.

mode 02

Improving an existing product

I join your project: new features, integrations, optimization, fixing and untangling existing code.

mode 03

Consulting & audit

Help choosing a stack, scoping a task, designing architecture or figuring out AI and automation.

~/faq $ man collaboration

Frequently asked

How much does it cost?

It depends on scope and complexity. After discussing the task and spec I give a fixed estimate or rate — no surprises along the way.

What about timelines?

A landing page — from a few days; a web app — from a couple of weeks. Exact timelines are locked at the estimate stage.

What happens after launch?

I hand over the project with documentation. Optionally I keep supporting and extending it.

~/contact $ ./contact.sh

Tell me about your task

Describe in a couple of lines what you need — I’ll reply and suggest the best way to build it. Usually in touch within a day.