FIELD PROOF
Results measured
in production.
UNNET supports the full path from assessment and PoC to migration, stabilization and long-term operations.
Retail Company
Migrated four core systems and reduced a critical batch from four hours to two.
Public Metro company
Migrated and validated approximately 8.2 billion rows across five databases.
Korea’s 1st ATS (Alternative Trading System) Company
Standardized operational databases and built an eight-server, 150TB Lakehouse.
PostgreSQL tuning
Improved critical SQL, throughput and CPU efficiency with evidence-based tuning.
7 cases
01
FINANCE / 2024–2026
Enterprise-wide standard DB and Lakehouse for Korea’s first alternative trading system
Standardized website, member-company, SOR and internal business databases on TarantulaDB and deployed a 150TB Lakehouse for an alternative trading system in Korea.
150TB data platform- Stable operations since March 2025
- 120GB+ daily data growth
- Eight servers and a 150TB Lakehouse
- One support team for databases and analytics
02
RETAIL / 2026.01
Core-system database migration for a major retailer
Migrated four core systems from overseas commercial PostgreSQL to TarantulaDB and reduced a critical batch process from four hours to two.
50% shorter batch processing- Migrated four core systems, application servers and operating systems together
- Incident-free migration in approximately four months
- Critical batch: four hours → two hours
- Stable operations after go-live
03
FINANCE & AI / IN PROGRESS
AI Vector DB expansion at a securities company
Expand an existing TarantulaDB environment into on-premises RAG to search internal documents, research and consultation knowledge.
Expand from operational DB to RAG- Connect Tarantula AI Vector DB with LLM/RAG
- Retain sensitive data on premises
- Use existing PostgreSQL operating practices
- Single-vendor technical support
04
FINANCE / 2025–2026
AML and PowerBase inquiry services for a capital-markets IT specialist
Following validation for anti-money-laundering systems, TarantulaDB was adopted again for PowerBase customer-facing inquiry services.
Successive adoption for critical services- AML deployment in October 2025
- PowerBase inquiry-layer deployment in March 2026
- Minimize changes to existing SQL assets
- Optimize CDC, ETL and high-volume inquiries
05
PUBLIC / 2026.04
Large-scale database migration for a public transport operator
Migrated approximately 8.2 billion records from commercial databases across five different systems to TarantulaDB and validated complete data consistency.
8.2 billion records migrated without data loss- Unified migration across five databases
- Approximately 8.2 billion records transferred
- Completed in approximately three months
- Parallel extraction of BLOBs and large partitions
06
CLOUD · PERFORMANCE / INCIDENT ANALYSIS
Cloud PostgreSQL incident resolution for a home-shopping company
Analyzed mobile-app latency on managed cloud PostgreSQL and improved performance step by step through indexes, Vacuum, pooling and reader distribution.
Average response: 140ms → 32ms- Average query response: 140ms → 32ms
- Mobile-app latency root-cause analysis
- Eliminate Vacuum delays
- Pooling and reader-distribution guidance
07
RETAIL · PERFORMANCE / FIVE-DAY ONSITE ASSESSMENT
PostgreSQL performance optimization for a home-shopping company
Improved indexes, SQL, Vacuum and partitioning in stages to significantly accelerate key product queries and overall processing.
Product query: 12 seconds → 7ms- Product-query SQL: 12 seconds → 7ms
- TPS 1,000 → 2,000+
- CPU utilization: 50% → 25%
- 82 index optimizations
NO MATCHING CASES
No cases match these filters yet.
Choose another category or return to all cases.
DELIVERY MODEL