AI Briefing
KO

Sajo Systems' Journey to Building Agentic AI Based on In-House ERP Data – Using the Strands Agents SDK and Kiro

·2026.07.05 20:52

Key point

Sajo Systems built an Agentic AI that analyzes ERP data and provides insights using the Strands Agents SDK and Kiro.

1 / 2

Details

To upgrade its existing groupware chatbot SAJO AI, Sajo Systems carried out an Agentic AI build project that goes beyond simple RAG to analyze data from a specific service and derive insights.

For this, they built a Spring Boot-based Java API Server to create an environment that can access journal entry (accounting ledger) data within the ERP. To improve data efficiency, they carried out column abstraction work, and for security, they applied a JWT Token authentication scheme.

During development, they applied the SDD (Spec Driven Development) approach using Kiro to maintain alignment between requirement specifications and tasks. They also designed agents capable of performing complex tasks by leveraging the Multi-Agent pattern of the Strands Agents SDK.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.