| NIM | H1D022058 |
|---|
| Namamhs | NABILLA TSANI AYASI |
|---|
| Judul Artikel | INTEGRASI LARGE LANGUAGE MODEL DAN RETRIEVAL-AUGMENTED GENERATION GEMINI SEBAGAI AI ASSISTANT SISTEM MANAJEMEN PROYEK TALL STACK |
|---|
| Abstrak (Bhs. Indonesia) | Manajemen proyek menghadapi tantangan kompleksitas data dan kebutuhan interaksi cerdas di era digital, di mana sistem konvensional kurang adaptif terhadap informasi real-time dan Large Language Model (LLM) murni mengalami halusinasi. Penelitian ini bertujuan mengembangkan AI Assistant berbasis Retrieval-Augmented Generation (RAG) dengan Gemini API pada sistem manajemen proyek TALL Stack (Tailwind CSS, Alpine.js, Laravel, Livewire) untuk meningkatkan interaktivitas, akurasi, dan produktivitas. Metode penelitian menggunakan pendekatan prototyping, meliputi identifikasi kebutuhan, desain (flowchart, DFD level 0-3, ERD, PDM), implementasi prototype web fungsional, refinement, dan evaluasi melalui black box testing, metrik Precision/Recall/F1-score, serta user-based evaluation via kuesioner kepada staf CV Jenderal Solusi Digital. Objek penelitian adalah sistem manajemen proyek berbasis web dengan fitur AI Assistant (chatbot, pembuatan proposal/fitur/laporan otomatis). Hasil menunjukkan sistem berhasil diimplementasikan dengan akurasi respons AI tinggi (Precision 98,75%, Recall 97,81% F1-score 98%), seluruh fungsi black box testing valid 100%, dan kepuasan pengguna 4,13/5 melalui user-based evaluation. Selain itu, sistem menunjukkan efisiensi penggunaan melalui otomatisasi berbasis konteks database real-time. Integrasi LLM-RAG Gemini pada TALL Stack terbukti efektif mendukung transformasi digital manajemen proyek dengan interaktivitas superior dan akurasi terverifikasi. |
|---|
| Abtrak (Bhs. Inggris) | Project management faces challenges related to data complexity and the need for intelligent interaction in the digital era, where conventional systems are less adaptive to real-time information and pure Large Language Models (LLMs) are prone to hallucinations. This study aims to develop an AI Assistant based on Retrieval-Augmented Generation (RAG) using the Gemini API within a TALL Stack (Tailwind CSS, Alpine.js, Laravel, Livewire) project management system to enhance interactivity, accuracy, and productivity. The research method adopts a prototyping approach, encompassing requirements identification, design (flowcharts, DFD levels 0–3, ERD, PDM), implementation of a functional web prototype, refinement, and evaluation through black box testing, Precision/Recall/F1-score metrics, and user-based evaluation via questionnaires administered to staff of CV Jenderal Solusi Digital. The research object is a web-based project management system featuring an AI Assistant (chatbot, automated proposal/feature/report generation). The results indicate that the system was successfully implemented, achieving high AI response accuracy (Precision 98.75%, Recall 97.81%, F1-score 98%), 100% valid results in black box testing, and a user satisfaction score of 4.13/5 from the user-based evaluation. In addition, the system demonstrates operational efficiency through automation based on real-time database context. The integration of Gemini LLM-RAG within the TALL Stack is proven effective in supporting the digital transformation of project management with superior interactivity and verified accuracy. |
|---|
| Kata kunci | Gemini, Large Language Model, Manajemen Proyek, Retrieval-augmented Generation, TALL Stack |
|---|
| Pembimbing 1 | Ir. Nofiyati, S.Kom., M.Kom., IPM |
|---|
| Pembimbing 2 | Muhammad Ihsan Fawzi, S.Kom., M.Kom. |
|---|
| Pembimbing 3 | |
|---|
| Tahun | 2026 |
|---|
| Jumlah Halaman | 176 |
|---|
| Tgl. Entri | 2026-01-27 09:48:49.958344 |
|---|