Makia Eni Timothy
AI & Software Engineer
Intelligent Web Apps,
Expertly Engineered.
I build modern web applications and design custom AI solutions, specializing in LLM integrations, Retrieval-Augmented Generation (RAG), and full-stack development.
Makia Eni Timothy
AI & Software Engineer
Intelligent Web Apps,
Expertly Engineered.
I build modern web applications and design custom AI solutions, specializing in LLM integrations, Retrieval-Augmented Generation (RAG), and full-stack development.
RAGBOT.AI
An educational RAG (Retrieval-Augmented Generation) assistant using Supabase PGVector, Llama 3.3, and Gemini embeddings for context-grounded Q&A.
“Eni delivered exceptional results that exceeded our expectations.”
RAGBOT.AI
An educational RAG (Retrieval-Augmented Generation) assistant using Supabase PGVector, Llama 3.3, and Gemini embeddings for context-grounded Q&A.
Project Overview
RAGBOT.AI is a custom Retrieval-Augmented Generation assistant built to answer student questions based on course syllabus materials. It uses vector search to retrieve relevant slide text and provides interactive citations.
Key Features
PGVector Search: Uses Supabase PostgreSQL with HNSW cosine similarity to match user queries to document slides.
Multimodal Document Ingestion: Integrates Groq Vision OCR to read text from uploaded images and documents.
Rate Limit Handling: Implements sequential ingestion processing with exponential backoffs to prevent API rate limits.
