Kamil Kerimov
Frontend Product Engineer

Kamil Kerimov

I build polished product interfaces, frontend systems, and Python-backed automation where motion, state, accessibility, and delivery discipline all matter.

Product UI Architecture Motion + Interaction Systems Python Automation Backends Telegram Product Workflows

This portfolio is also a shipped interface: responsive layout, theme runtime, adaptive effects, command navigation, CI evidence, and tested interaction flows. Read full profile.

Interactive Case Surface

UI state active
State Model Synced
Motion Budget Adaptive
Content Flow Case study
Proof Layer CI backed
  • 09:41:13 Hero surface presents frontend craft before infrastructure detail. ui
  • 09:40:57 Case study connects UI decisions to Python automation behavior. case
  • 09:40:22 Quality evidence stays available without taking over the first impression. proof
  • 09:39:41 Debug tools are hidden from visitors and can be opened with ?debug=1. dev
Flagship Case Study

Trendyol Price Tracker, shaped as a product experience.

The strongest project here is not only the bot. It is the full loop: user intent, product tracking, alert logic, history, admin visibility, and the interface thinking needed to make the workflow feel usable.

Problem Marketplace prices change faster than users can manually watch them
Product Loop Track, compare, alert, export, recommend, and manage subscriptions
Frontend Lens Every command is treated as a stateful user flow, not a one-off script
Engineering Proof SQLite WAL, locale validation, backups, readiness checks, and CI
Case Surface A product workflow model: capture intent, normalize data, explain state, and close the loop.
Source ingest and fetch 1.2s
Snapshot write + fallback 92ms
Rules + quiet hours 32ms
Telegram Dispatch 158ms

UI Flow: The Telegram surface is designed like a product UI: clear entry points, predictable states, helpful history, and commands that return users to the next useful action.

Data Pipeline Async scrape with fallback and timeout handling

Trendyol fetch paths cover direct URLs plus trending, search, and category scenarios.

Database Layer Migrations, indexes, WAL, PRAGMA, native backup

The SQLite layer is tuned for runtime use, not just local storage convenience.

Product Surface History, charts, compare, recommendations, quiet hours

The bot goes beyond price alerts into an actual retained product experience.

Operational Maturity Readiness 6/6, smoke checks, logs, and test contour

It shows deployment thinking and maintenance discipline, not just feature shipping.

Interface Lab

The site is part of the portfolio, not a wrapper around it.

I use this surface to show how I think about frontend craft: state, responsive density, motion, accessibility, and product storytelling working together instead of competing for attention.

01 Interaction Model

Command navigation, tabbed case views, adaptive hero state, keyboard-friendly controls, and mobile menu behavior are treated as product flows, not decorative scripts.

02 Motion Discipline

Effects are layered around device capability and reduced-motion settings, so polish can scale down before it becomes friction.

03 Proof-Aware UI

The UI gives visual weight to real proof: repositories, CI checks, Lighthouse, responsive testing, and concrete project behavior.

Design System Thinking

Visual polish is useful only when the system underneath is coherent.

The interface uses shared tokens, restrained glass surfaces, predictable controls, theme-aware color, and responsive constraints so the site feels designed rather than assembled.

Color

Light/dark tokens keep the same hierarchy while preserving contrast.

Theme-aware contrast and accent hierarchy
Type

Display copy stays bold; technical labels stay compact and scannable.

Display, body, and mono roles stay distinct
States

Controls expose state without changing layout dimensions.

Controls expose hover, focus, active, and mobile states
Layout

Grid tracks, card heights, and action rows are constrained for responsive stability.

Dense enough for product work, calm enough to scan
CI Evidence

Quality gate signals from real checks, not claims.

Build, type-check, unit tests, end-to-end flows, and Lighthouse are summarized from the latest snapshot generated in this repository.

Branch: pending Commit: pending Coverage: pending

Unit Tests

pending --
  • Passed: --
  • Failed: --

E2E Flows

pending --
  • Desktop: --
  • Mobile: --

Lighthouse

Perf --
A11y --
Best --
SEO --

Waiting for the latest quality snapshot.

Selected Projects

Other builds, same systems mindset.

Each project preview surfaces internal behavior, not only static screenshots.

01

Trendyol Tracker Core

Python | Aiogram | SQLite
Recommendations + Compare + Export WAL + Migrations + Backup 115 Tests / 40 Files + Green CI Contour

Telegram tracker with full product loop: subscriptions, diff-based alerts, price history, multilingual UX, and operational diagnostics tuned for stable real-world runs.

02

Blaster Game

Python 3.11 | Pygame
Quality Presets + Screen Shake Power-Ups + Telegraph Boss JSON Validation + CI + PyInstaller

Built with Python 3.11 + Pygame as a full 2D stack: real-time visual pipeline, procedural sound synthesis, multi-phase boss logic, replay buffer, and CI-backed Windows build flow.

03

Repository Workflow

GitHub | Delivery
Versioned Delivery Modular Codebases Iteration Ready

Additional repositories with practical automation tools, bot systems, and frontend builds.

Frontend Architecture Product UI Systems Responsive Interaction Motion Design Accessibility States Python Automation Telegram Products CI-backed Delivery
Build Together

Need a product interface that behaves like a real system?

Available for focused freelance builds and long-term product collaboration across frontend, automation, and bot workflows.

Frontend + Python Workflow Automation Production Quality