Arifian.AI
2025
About This Project
Rainfall Prediction with LSTM + Attention
This project explores the use of LSTM (Long Short-Term Memory) and LSTM with an Attention Layer to predict daily rainfall rates in Batam City. The primary goal is to assess the performance difference between the two models when the Attention Layer is implemented. The study leverages the BMKG DataOnline climate dataset and aims to improve prediction accuracy, especially in the context of dynamic and extreme weather changes.
- Python
- Jupyter
- TensorFlow
- Keras
- Scikit-learn
- Flask
Student Management System QA Playwright
Full-stack Student Management System paired with a comprehensive QA Automation Suite. Features include role-based access control, student management, grading, report cards, REST APIs, and automated UI/API testing using Playwright and Python with Page Object Model architecture and HTML reporting.
- Python
- Playwright
- Next.js
- TypeScript
- Express.js
- PostgreSQL
Elara : Personal Assistant
Autonomous AI Personal Assistant and RAG engineering engine built with FastAPI (Python 3.11+) and a Hybrid RAG Pipeline fusing dense Gemini Embedding 2 (768d) with Supabase PostgreSQL pgvector 0.8.2 and Full-Text Search (FTS GIN) via Reciprocal Rank Fusion (RRF), reranked by Gemma 4 26B (`gemma-4-26b-a4b-it`) and generated by Groq GPT-OSS 120B. Features a 6-step conversational state machine for project intake and lead qualification with automated persistence to Supabase and real-time outbound alerts to Arifian's Telegram DM / Hermes Agent, along with a built-in Single Page Admin Console (Minimalist Brutalism) for knowledge base management, GitHub repo ingestion, dynamic system prompts, and latency monitoring.
- Python
- FastAPI
- Google Gemini Embeddings & Gemma
- Supabase
- PostgreSQL pgvector
- Cloudflare R2
- Telegram Bot API
- Next.js