AK.
Contact

AndriiKhaliavkin

I turn ideas into useful digital products, combining product thinking, quality engineering, and practical experimentation with AI.

Andrii Khaliavkin in a dark suit

01 Selected project — 2026

OCCAM PM / A small anti-Jira

A smaller
anti-Jira.

A project workspace for teams who want to manage the work, not a process around it. It starts with TODO and DONE, then adds depth only when it helps.

For PMs

Paste a rough outline and turn it into a nested plan with bulk create.

For developers

Connect your own AI client through MCP to create, search, and update work.

Open OCCAM PM TODO → DONE / MCP

Product evidenceBulk → hierarchy → QA

OCCAM PM bulk-create panel previewing a nested project plan
Bulk create: Start with an outline, not a blank project board.
OCCAM PM showing an expanded nested task hierarchy
Hierarchy: Keep the structure visible when the work gets deeper.
OCCAM PM filtered QA task view
QA view: Filter the work you need without adding another process layer.

02 Selected project — 2026

SQLDojo / SQL interview practice

Learn SQL
by querying.

A browser-based SQL practice environment with guided exercises, filtered challenges, interactive database maps, and a live playground for trying out your own ideas.

For learners

Move from SQL concepts to working queries against real database structures.

For interviews

Generate a focused challenge, inspect the schema, and check the result against a reference answer.

Open SQLDojo View GitHub
120 exercises / SQLite

Product evidenceSchema → solve → sandbox

SQLDojo schema map with the Track-to-Album relationship highlighted
Inspect the schema: Trace a relationship before writing the query.
SQLDojo challenge solution returning customers with no orders and a passing result
Verify the answer: Run the query, then check it against a reference result.
SQLDojo playground creating and querying a study notes table
Use the sandbox: Keep a live database session for experiments of your own.

03 Selected project — 2026

RF Practice / Robot Framework learning

Build skills.
Prepare for RFCP.

Learn Robot Framework in two ways: work through runnable exercises, then prepare for RFCP certification with 500+ questions organised around the syllabus.

Exercises

Write and run Robot Framework code in the browser, then inspect terminal output, logs, and reports.

RFCP prep

Choose a chapter or topic for focused practice, or take a timed 40-question exam and review missed answers.

Open RF Practice 500+ questions / Robot Framework

Product evidenceCode → exam → review

RF Practice exercise editor with a passing Robot Framework run in the terminal
Run the exercise: Write the test, see the checks pass, then inspect the generated artifacts.
RF Practice timed exam with 20 of 40 questions answered
Take the exam: Work through 40 questions with the timer and progress visible.
RF Practice exam result with a score and explanations for incorrect answers
Review the exam: Finish a timed exam, then revisit each wrong answer with its explanation.

04 Selected project — 2026

Fundacja Dzieci Wdowy / Client site

Public site
for the foundation.

Public website for Fundacja Dzieci Wdowy Józefa Grodzkiego, replacing their previous WordPress site.

Scope

About, projects, activities, bank-transfer details, and a contact form.

Stack

Next.js on Vercel. Contact mail through Resend.

Open the site Next.js / Vercel

Product evidenceLanding → donate → contact

Fundacja Dzieci Wdowy homepage with the foundation logo, mission, and donate button
Landing: The public homepage: what the foundation is, and how to donate.
Fundacja Dzieci Wdowy bank-transfer card with IBANs for five currencies
Donate: Bank-transfer details for PLN, EUR, GBP, USD, and CHF.
Fundacja Dzieci Wdowy contact form with name, email, and message fields
Contact: A contact form that sends mail to the foundation.

05 Selected project — 2026

AI-driven coop quest / In development

A gamified
coop quest.

A browser world for a cooperative, AI-driven quest. It is still in development: join a ship, walk the map, and see other visitors in real time.

Architecture

Unity WebGL in the browser. Cloudflare R2 and a Worker serve the build. A Node WebSocket world-server on Railway handles join and presence.

Keywords

2.5D ship map, WebGL, Workers, R2, WebSocket presence. Quest logic is not finished.

Open the prototype Unity Web / in development

Product evidenceCafeteria → weapons → AI lab

Cafeteria on the ship with four round tables and the player in the hall
Cafeteria: The hall: tables, walk routes, and the rooms around it.
Player in the weapons room with a missile rack, command chair, and cargo crates
Weapons: Inside the weapons room: station, loader, and stored ordnance.
Player in the AI lab with containment cans and a token monument
AI lab: The room with the AI cans.