As a driven and enthusiastic Technology Consultant specializing in AI/ML, I relish the challenge of transforming complex issues into cutting-edge solutions and continuously strive to bring innovation to the table.
(Automatically updated daily by GitHub Actions.)
Click a year to expand/collapse. Showing description, language, stars, forks, and topics.
2026 — 10 repos
- LLMTradingAgents — LLMTradingAgents is an automated trading system that orchestrates competition between different Large Language Models (LLMs). It uses a 3-agent architecture to analyze market data, propose trades, and validate risk. •
Python• ★ 1 • 🍴 0 • Topics: ai-agents, cryptocurrency, finance, langchain, llm, marketing-analytics, stock, trading, trading-algorithms, vite - nashville-traffic-stop-bias-analysis — End-to-end analysis of racial bias in Nashville Police Department traffic stops using interpretable machine learning, statistical tests, and fairness audits •
Jupyter Notebook• ★ 1 • 🍴 0 - artist-popularity-m3u8-playlist-generator — Pipeline that builds two 1500-track M3U8 playlists (popularity-first + deep cuts) from an ordered artist taxonomy. Aggregates iTunes & Deezer data, ranks tracks, enforces zero-overlap, validates structure, and supports precise per-artist quota control via CSV with priority-based allocation logic •
Python• ★ 0 • 🍴 0 • Topics: deezer, itunes, m3u8, m3u8-playlist, music, playlist, song, youtube, youtube-music, youtube-playlist - auto-research-finance — Autonomous, auditable quantitative finance agents orchestrated via Claude Code. Implements MCP-based tooling, multi-agent trading debates, arXiv-to-factor pipelines for reproducible financial research, and comprehensive financial report. •
Python• ★ 0 • 🍴 1 - chrome-llm-automation — Chrome Extension for agentic browser automation. Runs entirely in the Side Panel using Set-of-Marks (SoM) grounding for reliable, high-precision observation and action. Powered by LLMs your choice. •
TypeScript• ★ 0 • 🍴 0 - dns-whitelist •
Unknown• ★ 0 • 🍴 0 - Dyslexia-Web-Reader — Help dyslexia struggle via Chrome Extension (Manifest V3) that converts web articles into a distraction-free reader view with adjustable typography, guided reading, text-to-speech, and optional LLM-powered simplification. •
TypeScript• ★ 0 • 🍴 1 - HT-VideoGraph — HT-VideoGraph is the first Hierarchical Temporal Knowledge Graph (HTKG) for long video understanding. It addresses the fundamental limitations of current video RAG methods by modeling videos at multiple temporal granularities with multi-modal feature preservation. •
Python• ★ 0 • 🍴 0 - llm-wiki-bench •
Python• ★ 0 • 🍴 0 - whatsapp-ai-recall — WAIR: A privacy-first, local-only AI assistant for WhatsApp. Indexes chat history into an encrypted SQLite database using hybrid RAG (vector + full-text) and local LLMs (Ollama). Zero data leaves your machine. •
TypeScript• ★ 0 • 🍴 0
2025 — 13 repos
- AeroFighters-SNES-reinforce-learning — Reinforce Learning AeroFighter Snes game using Gym Retro •
Python• ★ 0 • 🍴 0 • Topics: arcade, dqn, game, gym, ppo, reinforcement-learning, retro, snes - AgenticReqToDesign — Agentic AI transforming Business Requirements into Technical Detailed Design •
HTML• ★ 0 • 🍴 0 • Topics: agentic-ai, ai, artificial-intelligence, autogen, gemini - bullet-hell-100kbbh-rl — Gymnasium env + DQN/Double-DQN to dodge bullets in 100KBBH. Live preview, video recording, TensorBoard, CSV logs, and HTML reports. •
Python• ★ 0 • 🍴 0 • Topics: bullet-hell, dqn, game, pytorch, reinfocement-learning - CanvasAI — Interactive 2D game development sandbox that allows you to create and modify games using the power of Google's Gemini AI •
TypeScript• ★ 0 • 🍴 0 • Topics: 2d-game, ai, artificial-intelligence, game, game-development, gemini - chatgpt-from-scratch — A comprehensive, zero-dependency notebook to build a modern GPT from scratch. Implements Llama 3 architecture (RMSNorm, RoPE, SwiGLU) and trains on FineWeb-Edu using PyTorch. Educational, clean, and hackable. •
Jupyter Notebook• ★ 0 • 🍴 0 • Topics: bigram, chatgpt, colab-notebook, llama3, llm, notebook, tokenization, transformer - gemini-chess — A minimalist chess simulator powered by Google Gemini. This project uses the Gemini API to evaluate board states and generate moves via FEN strings. Featuring legal move validation, it provides a clean environment for testing LLM strategic reasoning and spatial awareness in a classic game setting. •
TypeScript• ★ 0 • 🍴 0 • Topics: gemini, generative-ai, llm - interactive-mcq-trainer — A simple, elegant MCQ practice UI built with pure HTML, CSS, and JavaScript. Auto-detects single/multiple answers and highlights correctness instantly. Includes explanations, scoring summary, and review mode. •
JavaScript• ★ 0 • 🍴 0 • Topics: exam, mcq, mcq-test, multiple-choice, notebooklm, quiz, test - mediapipe-hand-recognition — Hand recognition system using MediaPipe for real-time hand tracking, static hand-sign classification, dynamic movement classification, and data collection for training custom ML models. •
Jupyter Notebook• ★ 0 • 🍴 1 • Topics: handgesture-recognition, machine-learning, mediapipe, ml, supervised-learning - mini-rl-zoo — Educational RL framework with Q-learning, SARSA, DQN, A2C, and PPO implementations. Features multi-seed experimentation, automated analytics, and visualization for algorithm comparison. •
Python• ★ 0 • 🍴 0 • Topics: cartpole, dqn, frozenlake, lunar-lander, pusher, q-learning, reinforcement-learning - pong-ai — A modern arcade classic fused with generative AI. Battle a digital opponent as Google Gemini (via Genkit) dynamically mutates the game after every point you score. Experience unpredictable shifts in gravity, physics, and scale in real-time. Where retro gameplay meets alchemical AI transformation. •
TypeScript• ★ 0 • 🍴 0 • Topics: arcade, gemini, generative-ai, llm, pong - T-ECD-Sequential-Recommender — Sequential Recommender (SASRec/GRU) benchmarking the T-ECD dataset (9.2M events). Features interactive testing dashboards, 2D/3D cross-domain embedding visualizations, and impact analysis. •
Jupyter Notebook• ★ 0 • 🍴 0 • Topics: embeddings, gru, gru4rec, notebook, pytorch, recommendation-system, recommender-system, rnn, sasrec, transformer - web-image-gallery — A lightweight web application to extract all image URLs from a given website and display them in a dynamic, responsive gallery with infinite scrolling •
JavaScript• ★ 0 • 🍴 0 - wildlife-camera-trap-ai-system — AI-powered wildlife monitoring system powered by YOLOv8, with real-time detection and automatic video recording on the edge device. Metadata and thumbnails sync to a web dashboard for easy browsing and filtering. •
Jupyter Notebook• ★ 0 • 🍴 0 • Topics: ai, animal, bird, camera, computer-vision, edge, onnx, wildlife, yolo
2024 — 3 repos
- ChatGPT-custom-instruction — personal experiments on ChatGPT custom instruction •
Unknown• ★ 2 • 🍴 0 - GeminiChatFile — Chat File with Gemini •
Python• ★ 1 • 🍴 9 • Topics: gemini, gemini-pro, llm - PollinationAI — AI Image Generation https://pollinationsai.netlify.app/ •
Unknown• ★ 1 • 🍴 0
2023 — 7 repos
- NewsFront — Classic News WebApp •
JavaScript• ★ 2 • 🍴 0 • Topics: news, newsapi, newsapp - NatBrian •
Unknown• ★ 0 • 🍴 0 - scrollscape — Endless Photos Scrolling •
JavaScript• ★ 0 • 🍴 0 - simple-digital-business-card — Deploy digital business card in Firebase •
CSS• ★ 0 • 🍴 0 - trading-simulation — A high-throughput trading simulation for calculating OHLC summaries from large transaction volumes. Built with a Go app server, it leverages Kafka for asynchronous message queuing and Redis for fast data persistence. Features include Protobuf for data serialization, Dockerized infrastructure, and an API for bulk transaction. •
Go• ★ 0 • 🍴 0 - Ubuntu-Theme •
JavaScript• ★ 0 • 🍴 0 - whatsapp-chatgpt •
Python• ★ 0 • 🍴 0
2022 — 1 repos
- DecoratorDesignPattern — Example of Decorator Design Pattern In Go •
Go• ★ 0 • 🍴 0
2021 — 1 repos
- Python-For-Data-Science-Essential-Training-Part-1 — Linkedin Learning https://www.linkedin.com/learning/python-for-data-science-essential-training-part-1 •
Jupyter Notebook• ★ 0 • 🍴 0
2020 — 7 repos
- campus_indoor_map — Interactive University of Maryland campus indoor map with mapbox and openstreetmap •
JavaScript• ★ 0 • 🍴 0 • Topics: mapbox, openstreetmap - EM-algorithm-implementation — Implement Expectation-Maximization algorithm •
Jupyter Notebook• ★ 0 • 🍴 0 - GameSalad — Spanish Grammar Racing Game •
Unknown• ★ 0 • 🍴 0 • Topics: game - Github-IO-Template-academicpages — Nathanael Brian Github Page •
JavaScript• ★ 0 • 🍴 0 - SG-Data-Structure — Snippet code of S-G-Tree and S-G-KD-Tree data structure •
Java• ★ 0 • 🍴 0 - TwoHatsGame — Simulation game, where the user set up a hat color for each of the two people, either Blue or Red. If at least one of the two is right, they both win. They are allowed to come up with a plan in advance, but the adversary who is deciding which hats to put on them will hear the plan and will actively work against it. •
Java• ★ 0 • 🍴 0 - VideoPlayer — Video player with extra features, such as time stamp labeling •
JavaScript• ★ 0 • 🍴 0
2019 — 2 repos
- Data-Science-Tutorial — Data Science pipeline tutorial •
HTML• ★ 0 • 🍴 0 - JapaneseTranslationDiagflow — Japanese Translation Chat Bot (with voice recorder). Translation hasn't been implemented yet. •
Java• ★ 0 • 🍴 0
Last updated: 2026-04-29T02:04:41.865Z