From diagnosing devices to building the software that runs on them.
I’m Oleksandr — also Alex. I build applications that connect interfaces, data, and automation, and I know computers from the hardware up. I’m finishing a Bachelor of Technology in Computer Systems at City Tech in December 2026.
Next.js · TypeScript · Supabase · OpenAI Whisper · Servlets · JDBC · C# · Windows Forms · ADO.NET
Operations
Hardware diagnostics and repair, OS troubleshooting, reimaging, log analysis, and technical runbooks. Windows, macOS, and Linux; networking coursework in TCP/IP, DNS, and DHCP.
Education
Bachelor of Technology in Computer Systems New York City College of Technology (City Tech), CUNY, expected December 2026.
Coursework
Operating systems, computer security, systems analysis and design, and computer systems management and support.
IT Support Specialist · Jun 2025 — Present · Brooklyn, NY
Hands-on support across the device lifecycle: intake, diagnostic triage, repair, reimaging, and quality checks.
Process 500+ devices per week within a 12-person managed services team supporting approximately 30,000 client devices across two offices.
Diagnose Windows and macOS hardware and operating system failures through log analysis and internal diagnostic tools.
Diagnose and repair Chromebooks and ChromeCarts, install and reimage systems, support end users, and write troubleshooting runbooks.
Experience
Outlier AI
AI Code Evaluator · Contract · Mar 2024 — Dec 2024 · Remote
Reviewed AI-generated code for correctness, edge cases, and how well it followed the task.
Evaluated approximately 20 coding tasks per week across Java, Python, C#, and SQL, including query design, unit tests, and technical answers; maintained a 97% quality rating.
Promoted to reviewer, auditing peer evaluations and revising prompts to improve output quality.
Experience
Dance Zone
Coach & Studio Manager · Aug 2019 — Present · Brooklyn, NY
A different kind of systems work: keeping a studio running and helping people learn.
Coach classes of 10–12 students while managing scheduling, billing, cash reconciliation, facility safety, and client communication.
Helped retain a core client base for over five years.
Experience
In practice
Follow the failure to its cause.
At CITADEL NY, I traced recurring Chromebook reimaging failures to missing encoding parameters. I worked with Lenovo engineering to correct the issue and restore reliable system-image writes.
That same approach carries into my software: understand the dependency, fix the cause, and document what works.
The running app playing a finished render, with word-by-word caption highlighting. Silent; assembled from successive browser captures.
Audiogram Engine
Transcript review
Whisper’s transcript, with low-confidence words underlined for correction.
Audiogram Engine
Align to source text
Paste the script that was read aloud and the transcript is re-timed against it locally.
Audiogram Engine
Styling
Title and thumbnail styling beside a live preview of the burned-in captions for each aspect ratio.
Audiogram Engine
Rendering
Both formats render together, with staged progress and a cancel control.
Audiogram Engine
Output
Rendered output · 9:16. The portrait MP4 exactly as the app rendered it. The sample narration is synthetic speech.Rendered output · 16:9. The landscape MP4 from the same render, with its generated thumbnail as the poster.
Audiogram Engine
The idea
One recording, a set of deliverables.
Sharing audio as video takes a transcript, readable captions, the right aspect ratio, and supporting assets. I built a React interface and Flask backend around a Python and Whisper pipeline so that work happens in one place: upload, transcribe and review, style, render.
Tailoring a résumé means choosing relevant experience and explaining it for a particular role. AI can help with the wording, but it can also introduce a responsibility, technology, or result that was never part of the original experience.
Veritas
What I’m building
A link between the statement and its source.
Verified profile the person’s actual experience.
Role context what is relevant to the opportunity.
Generated statement wording connected to supporting profile data.
Human review checking the meaning and accuracy of the result.
Veritas
Direction
Source data is part of the workflow.
Next.js and TypeScript for the web application, Supabase and PostgreSQL for the data layer. A source link helps someone review a claim; it does not by itself prove that the generated wording is accurate.
Veritas
Status
In development.
The current focus is linking generated content to profile data and keeping that relationship reviewable as the application evolves.
Java servlet handles the request and application logic.
JDBC + MySQL persist and query relational records.
HTML response the result, back in the browser.
Java Web Applications
Examples
Two entry points into the code.
Widget Inventory Manager — a servlet application where a factory assembles widgets and JDBC stores them in MySQL. Source ↗ (opens in a new tab)
IT Asset Console — validates technician credentials with a prepared statement and looks up workstation assignments. Source ↗ (opens in a new tab)
Java Web Applications
Scope
Learning applications with visible limits.
Authentication varies across the applications; the public inventory example uses a username greeting, not enforced login. Stronger sessions, parameterized queries throughout, and integration tests are the next steps.
