Skip to content
View StepanTita's full-sized avatar

Block or report StepanTita

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
StepanTita/README.md

🌍 Hello, world! And greetings, fellow coder or recruiter! πŸ‘‹

Welcome aboard my GitHub time machine! πŸ•°οΈβœ¨ Settle in, and let me give you a tour of my code journey. Grab some β˜•οΈ and enjoy the ride!

Stepans's Github Stats

Top Langs

πŸ‡ΊπŸ‡¦ Origin: Kharkiv, Ukraine

This is where it all began...

  • πŸ–₯️ My first computer
  • πŸ‘¨β€πŸ’» My first line of code and repository
  • πŸ› My first bug
  • πŸ’Ύ First commit & pull request
  • πŸŽ“ First bachelor's degree in Software Engineering from Kharkiv National University of Radio Electronics

NURE

πŸ—½ Current Location: New York, USA

I've made my way here for an upgrade!

  • πŸŽ“ Earned my Master's in Computer Science (Data Science concentration) at Fordham University β€” 4.0/4.0 GPA. Thesis: "Context Attribution To Attain Generalizability in Non-fine-tuned Transformer...", advised by Dr. Ruhul Amin.
  • πŸ’Ό Now working as a Senior Applied Machine Learning Engineer at Labelbox (remote), building on prior roles as a Generative AI & Machine Learning Engineer at adMarketplace and a Graduate Researcher at Fordham.

Fordham

🌌 Exploring My GitHub Universe

My GitHub captures milestones, learnings, and challenges. It's more than code; it's a snapshot of my journey. πŸš€

  • πŸ† 2019: Before the code got serious, there were competitions - Social Media Award at the International Startup World Cup (Denmark) with the sign-language translation app "Sounce", 1st place at an AI Hackathon at NURE's Scientific Park "Synergy" (plus a solo trip to Turkey for best presentation πŸ‡ΉπŸ‡·), and a Microsoft Student Partners Ukraine hackathon win with a fake/spam Instagram detector - inst_custom_api.

  • πŸ“… 2019: My first end-to-end project with UI/UX, backend, algorithms, and visualizations - bstree-tango.

  • πŸ•°οΈ Pandemic Era (2020): Developed a PWA to help me (and others) keep track of sporadic online classes in Spain - pwa-timetable.

  • πŸ“… 2021: Bachelor's thesis (High Honors, 97.74/100) representing a smart drone system for road quality assurance - rdrone (frontend).

    • Research paper on fine-tuning transformers for Ukrainian Language news-contest
  • πŸ“… 2022: Dive into my research projects - q-learning & semi-supervised-regression.

    • And me having fun training LLM on my telegram chats with a mix of anecdotes and math tasks πŸ‘€ llm-friend-bot
  • πŸ” Crypto Projects (2023): Current explorations include crypto-news & crypto-trading.

  • πŸ”¬ Research aspirations (2023-2024): My research focus currently falls on explainable and interpretable Language Models, including LLMs

    • Nano-BERT & space-model, the latter earning a Best Paper Award, a Spotlight at AGI Leap Summit 2024, and People's Choice at Fordham-GSAS's 2024 Three Minute Thesis Competition πŸ†
    • CAM-BERT Context Attribution Language Model Fine-tuning Framework & CAM-Explain Context Attribution explanation benchmarking
  • πŸ› οΈ Building from scratch (2024-2025): Went down the "reimplement it yourself" rabbit hole - nanotorch, a from-scratch PyTorch reimplementation covering autograd, attention, LoRA, and quantization.

    • Also shipped a full-stack Warehouse Management System (Django + Vue + Kotlin + Raspberry Pi IoT) and a Go CLI for chatting with Bing Chat - go-EdgeGPT.
  • πŸ€– NeurIPS 2025: Co-authored the Labelbox Research team's reflections on the conference - "Reflections on NeurIPS 2025: Advancing Evaluation and Continual Learning in AI".

πŸ“‘ Papers & Conferences:

  • Multimodal News Source Classification with NLP Transformers (2022) [ICTERI, Code]
  • Space Model Framework for Conceptual Modelling in NLP (2023) [AAAI, ArXiv, Code]
    • πŸŽ‰ Best paper award
    • 🌟 Spotlight Presentation, AGI Leap Summit 2024
    • πŸ₯‡ People's Choice, Fordham-GSAS Three Minute Thesis (3MT) Competition 2024
  • Transformer Context Attribution for Fake News Detection (2024) [ICLR, ArXiv, Code]

πŸ”­ Focus & Interests

Deep Learning & Natural Language Processing (NLP) have captivated my heart and mind. The potential in these branches of machine learning is vast, and I'm all in!

🧰 Professional Toolkit

  • Languages: Python, GoLang, C++, SQL
  • Technologies: Databricks, PySpark, Hive, PyTorch, Tensorflow, PostgreSQL, NoSQL (Redis, Mongo)
  • Tools: gRPC, REST, Docker, Terraform, Bash, AWS, Google Cloud

🌟 A Bit More About Me

I bring nearly 2 years of experience in the crypto sphere, working as a backend developer, primarily with GoLang. I then transitioned into the CRM domain for a year, giving me a robust understanding of micro-service architectures. While I've delved into multiple languages professionally, Python remains my true love ❀️. The highlight of Summer 2023? My ML/DS internship with the Volvo Group. That was followed by a stretch as a Graduate Researcher at Fordham (alongside my Master's), then a role as a Generative AI & Machine Learning Engineer at adMarketplace. These days, I'm working on inspiring NLP and LLM projects as a Senior Applied Machine Learning Engineer at Labelbox.


I'd love to connect, collaborate, or just chat. πŸ’Œ Reach out if you're as passionate about tech and innovation as I am!


Follow Stepan on LinkedIn

Pinned Loading

  1. crypto-trading crypto-trading Public

    Crypto Arbitrage Explorer: A Dash-powered Python tool for real-time visualization of cryptocurrency arbitrage opportunities across exchanges.

    Jupyter Notebook 25 2

  2. go-EdgeGPT go-EdgeGPT Public

    GoBingChat: A GoLang-powered CLI tool to seamlessly interact with Bing Chat, enabling easy integration into any application.

    Go 7

  3. semi-supervised-regression semi-supervised-regression Public

    COREG with Transformers Exploration: This research project delves into enhancing the COREG algorithm by incorporating transformer models, particularly DistilBERT and DeBERTa. Through rigorous exper…

    Jupyter Notebook 5

  4. q-learning q-learning Public

    a Python-based platformer infused with Q-Learning and dynamic level creation from simple JSON files.

    Python 5

  5. news-contest news-contest Public

    This repository contains Jupyter notebooks detailing the experiments conducted in our research paper on Ukrainian news classification. We introduce a framework for simple classification dataset cre…

    Jupyter Notebook 4

  6. pacman pacman Public

    This is a revised python implementation of the original retro-game Pacman. It has some new ghosts and bonuses types, as well as specific gameplay traits. This is intended to be an environment for f…

    Python 4