Portfolio

A sample of automation, analytics and AI assistant engagements delivered for fintech, media and SaaS teams.

DemoWeb

Air Pillow

AirPillow is an interactive web page with a 3D air-quality globe, where pollution levels are shown using a color scale (Good / Moderate / USG / Unhealthy / Very Hazard). You can click on the globe or use the city/country search to highlight a location and quickly assess the current conditions.

Stack:Python · HTML · CSS · JavaScript

API:FastAPI (HTTP + WebSocket)

Integrations:Open-Meteo API · OpenAQ API

Outputs: Browser UI · MJPEG stream

Frontend: HTML · CSS · JavaScriptBackend: Python · FastAPI · UvicornHTTP Client: httpxStreaming: Playwright · WebSocket · MJPEGIntegrations: Open-Meteo API · OpenAQ APIConfig: python-dotenv

Read case study

DemoWeb

BuhUchetAI

BalanceCloud is an AI-powered accounting learning platform. It helps users practice bookkeeping and accounting tasks with instant AI feedback, progress tracking, and an AI chat assistant. The platform is designed for students and learners, not for enterprise accounting operations.

-Faster understanding of accounting concepts through hands-on practice Immediate AI feedback reduces learning friction Structured progress tracking improves learning consistency

Backend: Python · FastAPI · Uvicorn · Jinja2 · ReportLabDatabase: SQLite · SQLAlchemyFrontend: HTML · CSS · JavaScriptAuth: Passlib · bcryptIntegrations: OpenAI API · PayPal SDK

Read case study

DemoWeb

Cryptocurrency Analytics Dashboard with Telegram Bot Integration

Developed a real-time cryptocurrency analytics dashboard using Python, Dash (Plotly), and WebSocket API from Binance. Integrated a custom Telegram bot (aiogram) for instant chart updates and user control directly from Telegram. Users can select cryptocurrency pairs and time intervals, and instantly view live price charts with advanced indicators (EMA, SMA). Automated server-side synchronization and ensured a seamless UX between the web dashboard and the bot. The project demonstrates skills in API integration, data streaming, and modern Python back-end development.

-Delivered a real-time crypto monitoring experience with continuous WebSocket streaming (no manual refresh). Reduced time-to-insight by adding interactive candlestick charts with EMA/SMA overlays. Enabled Telegram-first control: users pick a pair and timeframe in the bot and instantly open/update the chart. Kept the experience consistent via server-side state sync between the dashboard and Telegram bot.

PythonDashPlotlyWebsocketsBot

Read case study

DemoWeb

Auralance

Auralance is a modular web platform that helps freelancers find, filter, and manage jobs efficiently. It features dynamic job filters, favorites, search history, and subscription notifications, all using modular HTML, CSS, and JavaScript. Designed for scalability and ease of maintenance, the responsive UI works locally without a backend, with optional integration for Flask or Django. The project emphasizes clean code and user-friendly experience.

-Auralance delivers a clean, conversion-friendly landing experience for a freelance job discovery product. The new UI simplifies searching with a clear filter flow (basic → advanced), improves content scannability with consistent job cards, and adds retention loops via Favorites, Search History, and Subscribe. The result is a more predictable user journey from “browse” to “shortlist” to “come back later”.

HTML5CSS 3JavaScriptResponsive DesignFront-End Development

Read case study

LiveAI Bot

AI_Trader FX & Crypto Outlook

Telegram bot delivering a daily scenario for FX and crypto by combining key news (macro, Fed/ECB, regulators, geopolitics) with live price/volatility data. Every hour, posts short-term BTCUSD and EURUSD projection charts. Also supports on-demand spot updates via buttons with concise, actionable signals.

FX (EURUSD), Crypto (BTCUSD)Markets

Daily brief + hourly chartsCadence

PythonFastAPITelegram Bot APICharting

Read case study

LiveSaaS

Analytics Command Center

Streaming metrics from marketing, product and finance sources into a single pane of glass with proactive alerts.

15 minTime-to-signal

6Teams onboarded

Next.jsClickHousedbtAWS Lambda

Read case study