Implementasi Sistem Analisis Kelayakan Kredit Sebagai Pra-Evaluasi Menggunakan Large Language Model Pada Workflow N8N

Ratnasari, Gita (2026) Implementasi Sistem Analisis Kelayakan Kredit Sebagai Pra-Evaluasi Menggunakan Large Language Model Pada Workflow N8N. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221049-Undergraduated_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Penyaluran kredit merupakan proses penting dalam industri perbankan karena berkaitan dengan pengendalian risiko gagal bayar dan ketepatan pengambilan keputusan.Penelitian ini bertujuan menerapkan Large Language Model (LLM) untuk menentukan kelayakan kredit berdasarkan profil calon debitur dengan memperhatikan ketentuan yang berlaku, serta menganalisis dan membandingkan kinerjanya dengan metode
klasifikasi konvensional. Metodologi penelitian meliputi studi literatur, pengumpulan dan prapemrosesan data, pembentukan Standar Operasional Prosedur (SOP), implementasi
AI workflow pada platform n8n, pembuatan sistem interface, serta verifikasi dan validasi. Dalam AI workflow, data kredit dari spreadsheet dianalisis menggunakan LLM Cohere
berdasarkan profil calon debitur dan SOP analisis kredit. Hasil analisis kemudian diatur menggunakan Structured Output Parser dan digunakan untuk memperbarui status data pada spreadsheet. Evaluasi dilakukan menggunakan confusion matrix
serta metrik accuracy, precision, recall, dan F1-score, kemudian dibandingkan dengan Logistic Regression, Random Forest, Support Vector Machine, Multilayer Perceptron
dan Tabnet. Hasil pengujian menunjukkan bahwa LLM memperoleh accuracy sebesar 93,68%, precision 95,40%, recall 91,21%, dan F1-score 93,26%, serta menunjukkan kinerja lebih tinggi dibandingkan metode konvensional pada penelitian ini. Dengan demikian,penerapan LLM melalui AI workflow n8n berpotensi mendukung analisis kelayakan kredit yang terstruktur, konsisten, dan informatif.
======================================================================================================================================
Credit approval represents a critical process in the banking industry, as it is closelytied to controlling default risk and ensuring accurate decision-making. This study seeks to apply a Large Language Model (LLM) to assess creditworthiness based on prospective debtor profiles, while adhering to prudential the provisions , and further aims
to evaluate and compare its performance against conventional classification methods. The methodology employed encompasses a literature review, data collection and preprocessing,
the formulation of Standard Operating Procedures (SOP), the implementation of an AI workflow on the n8n platform, the development of a system interface, and a verification
and validation process. Within this AI workflow, credit data drawn from a spreadsheet are processed by the Cohere LLM in reference to prospective debtor profiles and the
credit analysis SOP. The resulting analysis is then structured through a Structured Output Parser and applied to update the corresponding data status in the spreadsheet. Model evaluation relies on a confusion matrix together with the accuracy, precision, recall, and F1-score metrics, with the outcomes subsequently benchmarked against Logistic
Regression, Random Forest, Support Vector MachineMultilayer Perceptron and Tabnet.Findings indicate that the LLM attains an accuracy of 93.68%, precision of 95.40%,
recall of 91.21%, and F1-score of 93.26%, outperforming the conventional methods examined in this study. These results suggest that deploying an LLM through an n8n based AI workflow holds promise for enabling creditworthiness assessments that are more structured, consistent, and informative.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Analisis Kelayakan Kredit, Large Language Model, AI Workflow, n8n, Klasifikasi Kredit, Creditworthiness Analysis, Large Language Model, AI Workflow, n8n, Credit Classification.
Subjects:	Q Science > QA Mathematics Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Gita Ratnasari
Date Deposited:	31 Jul 2026 03:56
Last Modified:	31 Jul 2026 03:56
URI:	http://repository.its.ac.id/id/eprint/140778

Actions (login required)

View Item