The Forefront of Legal AI! Behind the Scenes of LegalOn Technologies' Google Cloud Migration and Practical Application

Hello. In this article, we will unravel the Google Cloud adoption case study of LegalOn Technologies, a rapidly growing startup in the legal domain.

How can generative AI and cloud technology be integrated into highly specialized fields like law and real estate to accelerate business? This case study is packed with insights, not just from a technical perspective, but also from the viewpoint of how to transform operational processes in the field. For those learning IT in their 30s or those looking to combine their professional expertise with technology, the architectural philosophy and practical AI application methods should serve as a great reference.

Consolidating Product Infrastructure on Google Cloud

The company develops services such as "LegalForce," an AI-powered contract review support service, and "LegalForce Cabinet," a contract management system.

Previously, these services operated on different cloud platforms depending on their purpose and use case. However, with the development of "LegalOn Cloud," the next-generation AI legal platform, they made the decision to consolidate their product infrastructure onto Google Cloud.

The deciding factors for this selection were as follows:

• The superiority of GKE (Google Kubernetes Engine) as a computing resource foundation

• High compatibility with existing data analysis infrastructure centered on BigQuery

• Google's status as a global leader in the AI field, which the company prioritizes

GKE Supporting Short-term Development and the Challenge of Scalability

The development period for "LegalOn Cloud" was extremely tight, at just one year.

To increase the speed of development and operations while coordinating numerous microservices, the company adopted the following approaches:

• They adopted GKE, which makes it easy to create a multi-tenant system that each development team can operate independently.

• They significantly reduced operational overhead by utilizing GKE Autopilot and Anthos Service Mesh.

• In the early stages of development, they prioritized speed and implemented strategies to transition deployment methods in phases.

Since the release, they have been working on large-scale refactoring of their infrastructure configuration with an eye toward future multi-region (overseas expansion) and multi-product strategies, continuing to evolve their system architecture in line with business growth speed.

Data Infrastructure and Metadata Strategy with Gemini

Google Cloud features are also being fully utilized in the area of data analysis.

• They have adopted Looker for BI tools and BigQuery for their data warehouse, promoting the creation of dashboards for various KPIs.

• For data users to effectively utilize dashboards, it is essential that they correctly understand the meaning of the data (metadata).

• To address this metadata management challenge, we are utilizing LLMs (Gemini) to automate the process. By using Gemini to generate dbt sources (such as descriptions) from schema information, we have dramatically reduced the cost of manual input.

Sales Transformation through Predictive AI x Generative AI

Beyond building a technical foundation, AI is also delivering results on the front lines of business in the sales domain. The company's case study was selected as a finalist in the '4th Generative AI Innovation Awards' hosted by Google Cloud.

• We used predictive AI to score the 'potential' of prospective customers.

• At the same time, we built a hybrid model that utilizes Gemini to generate the 'reasoning for that potential' in natural language and present it to sales representatives.

• Through this initiative, we have achieved concrete results, with the number of calls increasing by 16.8% and the conversion rate to business meetings improving by 15.1%.

Summary: The Fusion of Practical Work and Technology

The LegalOn Technologies case study demonstrates a highly multifaceted approach, ranging from the construction of cutting-edge cloud infrastructure (GKE and BigQuery) to diligent metadata maintenance using generative AI (Gemini), and direct AI application to sales activities.

[Ninja's Secret Postscript] We have further condensed and deepened the 'correct answers for practical application' seen from a thorough survey of 120 companies, including the case studies introduced in this article.

■ [Complete Preservation Edition] 30 Google Cloud Generative AI Use Cases We have carefully selected 30 companies out of 120 that are particularly 'effective for practical work right now.' We are currently releasing this premium-level analysis, which hits the critical points of architecture, at a special price.

Get your hands on the 'genuine wisdom' needed to survive in the wilderness.