Alif Kurniawan

LLM-based AI Agent Platform with RAG
Sole backend engineer on a 5-person product team (CTO, backend, frontend, QA, designer), owning the system
architecture end-to-end for an AI Agent platform powered by LLMs using LangChain and LangGraph, with RAG to
answer queries grounded in a knowledge base built from user-uploaded PDF documents.
Architected a decoupled backend: a FastAPI service handling client requests, and a separate worker service handling
LLM calls asynchronously, connected via RabbitMQ for reliable, scalable task processing.
Enabled multi-channel access to the AI Agent across 3 channels — WhatsApp, Instagram, and a website live chat
widget —with MongoDB as the primary store for conversation history and uploaded documents.
Containerized and deployed the API and worker as independent Docker services.
Leveraged AI-assisted development tools (Cursor, Antigravity) with iterative prompting to accelerate development of
complex agentic workflows and speed up debugging.
Other Backend & Mobile Projects

Built a Python-based crawling tool to systematically collect Twitter data, storing results in MongoDB for analysis and
archival.
Developed a RESTful API backend with FastAPI and integrated RabbitMQ as a task scheduler to orchestrate real-time
sentiment analysis on collected data.