# AgileSODA > AI\-powered document processing and machine learning operations for insurance and financial services ## At a glance | Field | Value | | --- | --- | | Category | Document Management | | Type | Solutions Provider | | Region | APAC | | Headquarters | Seoul, South Korea | | Founded | 2015 | | Stage | Established | | Employees | 51-200 | | Funding | Public | ## Overview AgileSODA is an enterprise AI software company based in Seoul, South Korea, providing machine learning operations (MLOps), intelligent document processing (IDP), and decision optimization solutions to carriers and financial institutions. The company serves the Life & Annuity and P&C insurance sectors with AI tools that automate document-heavy workflows such as medical expense receipt recognition, claims review, and underwriting data extraction. The company offers a suite of specialized AI platforms: SparklingSoDA handles end-to-end MLOps including model development, deployment, and monitoring; TwinReader delivers AI-based OCR and document understanding with GS Grade 1 certification from the Korea Information and Communications Technology Association; BakingSoDA applies reinforcement learning to optimize business decisions; and ChatSAM provides augmented search capabilities. These products are deployable on major cloud providers (AWS, Azure, GCP) and integrate into enterprise insurance workflows. AgileSODA is listed on the KOSDAQ exchange and has raised approximately USD 20.1M in total funding, with investors including KB Securities, Hanwha Investment & Securities, and BNK Securities. The company has documented deployments with Hanwha Life Insurance, Samsung Fire & Marine Insurance, and Hyundai Marine & Fire Insurance -- three of South Korea's largest insurance groups -- and has been recognized as an "AI + X TOP100" company for three consecutive years by the Korean government. ## Products & services ### SparklingSoDA Enterprise MLOps platform covering the full model lifecycle from development to deployment and ongoing monitoring. Version 4.0 introduced a ModelOps approach with continuous integration, continuous delivery, and continuous testing (CI/CD/CT) capabilities to maintain AI model performance over time. **Key features:** - Full pipeline management from model training through production deployment - Continuous monitoring and model performance tracking - CI/CD/CT framework for maintaining model integrity in production *Target users:* Insurance carriers and financial institutions deploying machine learning models at scale ### TwinReader AI-based Intelligent Document Processing (IDP) and OCR solution that extracts structured data from unstructured documents and images. TwinReader understands document semantics and converts handwritten or printed content into machine-readable formats. Version 2.2 includes advanced table recognition and holds GS Grade 1 certification. **Key features:** - AI OCR with semantic document understanding - Advanced table structure recognition - GS Grade 1 certification for quality and security compliance - Application in insurance medical expense receipt processing *Target users:* Insurance carriers automating claims intake and medical document processing ### BakingSoDA Reinforcement learning MLOps platform focused on optimization and automated decision-making workflows. Designed to apply RL-based techniques to complex business problems requiring adaptive decision logic. **Key features:** - Reinforcement learning model development and deployment - Business decision optimization workflows - Integration with enterprise data pipelines *Target users:* Carriers and financial institutions seeking AI\-driven process optimization ### ChatSAM Augmented search solution leveraging AI for enhanced enterprise information discovery and retrieval across internal knowledge bases and document repositories. **Key features:** - AI\-enhanced search across unstructured enterprise content - Natural language query support - Integration with document management systems *Target users:* Insurance operations and knowledge management teams ### TwinDoc NLP-based platform built on Transformer architecture for natural language processing and text understanding tasks, supporting document classification and content extraction workflows. **Key features:** - Transformer\-based NLP for text understanding - Document classification and content extraction - MLOps lifecycle support for NLP models *Target users:* Carriers and insurers requiring NLP\-based document intelligence ## Category & focus - Subcategories: AI OCR, Intelligent Document Processing, MLOps, Decision Optimization - Insurance verticals: Life & Annuity, P&C Commercial, P&C Personal - Target customers: Carriers ## Notable customers - Hanwha Life Insurance \-\- AI OCR for automated medical expense receipt recognition; Claims AI Automatic Review System - Samsung Fire & Marine Insurance \-\- Machine learning platform implementation - Hyundai Marine & Fire Insurance \-\- AI optimization solutions ## Links - Website: - Directory profile: - Linkedin: --- *Last updated: 2026\-06\-30*