This repository showcases a curated suite of AI projects representing the most in-demand technologies and capabilities for 2025. These tools address critical enterprise use cases, from developer productivity and code quality to retrieval-augmented generation (RAG), AI explainability, and local LLM deployment.
Each app is designed to be:
- 🔐 Privacy-first (fully local or hybrid cloud/local)
- 🧱 Modular & composable
- 📈 Enterprise-ready with extensibility in mind
An enterprise-grade Copilot alternative that offers:
- ✍️ Autocomplete for Python, JS, etc.
- 💬 Chat mode with memory
- 🧪 Unit test generation
- 🧱 Style & quality checks
- 🔁 Git auto-commit integration
- ☁️ OpenAI GPT-4 or ⚙️ Ollama (local models)
A fully offline Retrieval-Augmented Generation (RAG) system:
- 📄 Upload & chunk documents (PDFs)
- 🔍 Embed via
sentence-transformers - 🧠 Vector search with ChromaDB
- 🤖 Answer via local LLMs (Ollama)
Ideal for legal, finance, and research teams.
AI-powered static analysis tool with:
- 🔍 Code scoring
- 🧠 Refactoring suggestions
- 🚨 Smell detection & complexity checks
- 💾 Session persistence
- ☁️ GPT-4 or ⚙️ Ollama support
Fits perfectly into CI/CD or secure coding pipelines.
- ✅ RAG (Retrieval-Augmented Generation)
- ✅ LLM-powered code intelligence
- ✅ LLMs for QA, refactoring, and documentation
- ✅ Hybrid model support (cloud + local LLMs)
- ✅ Enterprise tooling: Git, sessions, audit trails
- ✅ Developer portals and explainability
- ✅ Offline-first architecture with full control
streamlit,openai,ollama,gitpython,pypercliplangchain,chromadb,sentence-transformers,pypdf- Compatible with:
- 🧠 Ollama models (Mistral, LLaMA2, Deepseek, Codellama)
- ☁️ OpenAI GPT-4 / GPT-3.5
# Clone the repo
git clone https://github.com/your-username/ai-2025-projects.git
cd ai-2025-projects/mini_copilot # or rag_local / code_quality
# Install dependencies
pip install -r requirements.txt
# Run the app
streamlit run copilot_app.py # adjust per app