wilson Ibekason@wilson_ibekason51

Code Compass – AI-Powered Open-Source Contribution Helper Code Compass makes open-source contribution easier by helping developers understand complex codebases, visualize dependencies, and find actionable contribution opportunities. Core Problem New developers struggle with large, unfamiliar repositories and lack clear starting points. Code Compass bridges this gap with AI-driven insights and interactive visualization. Key Features ✅ AI-Powered Analysis – GPT-5-based repository summarization, code explanations, and intelligent Q&A. ✅ Visual Knowledge Graphs – Interactive dependency maps using NetworkX + PyVis for files, classes, and functions. ✅ Smart Contribution Guidance – Detects TODOs, missing tests, and suggests beginner-friendly GitHub issues. ✅ Multi-Language Support – Handles Python, JavaScript, TypeScript, HTML, CSS, and config files. How It Works Repo Ingestion – Clone any GitHub repo via API. Graph Mapping – Build a knowledge graph of imports, function calls, and file relationships. AI Insight Layer – GPT-5 answers natural language questions, explains architecture, and recommends contributions. Interactive Exploration – Filter graphs, view file summaries, and discover improvement areas. Tech Stack Frontend: Streamlit, PyVis Backend: FastAPI, Python AI & NLP: GPT-5 (via AIML), Google Generative AI embeddings, FAISS for semantic search Visualization: NetworkX, Plotly Integration: GitHub API (PyGithub) Why It’s Innovative Unlike basic code search tools, Code Compass combines graph visualization, AI-driven explanations, and predictive contribution detection. It not only answers “what” and “where” but also explains “why” and “how”, empowering newcomers to contribute confidently. One-Line Summary: Code Compass transforms codebases into interactive, explainable maps, making open-source projects accessible to everyone.

Coherent Fox is an audit generator for Python code based on pre-trained BERT model. It is done to have a Powerful NLP Model based on Attention mechanism. We have build our own transformer from scratch. We have focused our training on a Python Code DataSet from Hugging Face. The idea is to provide a transformer that been built through main attention scores which are identifiers and separators from the study done on Java, from BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. We provide in the back-end Cohere API that will generate a response to a user prompt through our interface. We are a team of two full-stack devs with Experience in ML & Blockchain also.

🗓️ This will be a week of hacking and fun from 24 February to 3 March 💻 Create innovative new apps with OpenAI's latest AI tools 💡 Learn from top AI professionals ⚒️ Combine GPT-3, Codex, Dalle-2, and Whisper to build your AI app 🐱‍💻 Now is the time to register and let's get started!

🗓️ This 7-day virtual hackathon from January 27 to February 3 will be your time to shine! 💻 Construct AI apps with the latest LLM-powered technology from Cohere and try the newly released Multilingual Text Understanding model ✔️ There's no prerequisite level. Just starting out with AI? Or perhaps you're a veteran Data Scientist? A Designer? A Business Developer? We want you and your domain expertise! 🐱‍💻 Sign up now for free and let's get going!

🗓️ Take part in this 7-day virtual hackathon from March 10 to March 17! 💻 Create AI applications utilizing Cohere's LLM-powered Multilingual Text Understanding model and Qdrant's vector search engine. ✔️ Are you new to AI or an experienced data scientist? Designer, or business developer? Regardless of your experience and background, we welcome you and value your domain expertise. 🐱‍💻 Join us for free and let's get started!