Resona Asset Management announced on October 7, 2026, that it had built a system on Snowflake to automate the creation and revision of investment trust product summaries and prospectuses. Developed jointly with Nowcast, the system adds four functions to work previously handled by staff manually editing Word files for each fund: fund search, draft creation, bulk changes, and generation of comparison tables showing old and new text.

Changes to product summaries and prospectuses occur frequently when new funds are launched, internal rules are revised, or notices from groups such as the Investment Trusts Association, Japan, and changes in laws require updates. Responding to one revision notice previously took several days to about a week, and workloads grew with the number of funds. Storage of original Word files, naming conventions, and version management also depended on individual staff members.

The new system organizes a database of investment trust information and a folder structure in Resona AM’s existing Snowflake environment. It centrally manages information extracted from existing Word files as structured data. Users can filter fund searches by product category, investment region, and management style. They can also enter context in natural language, which AI converts into search criteria to identify similar funds for reference when creating documents or funds that need revisions.

A new-document and individual-change editor lets staff create product summaries or prospectuses based on similar funds, or revise documents for a specific fund. When staff enter change instructions, generative AI proposes draft wording that they can approve, discard, or edit directly on screen. The system can also export Word files with comments identifying the changed sections and their content.

For changes affecting multiple funds, AI drafts replacement rules pairing the original and revised wording based on revision notices or change requests. Software identifies the affected funds and scans their files; limiting AI to drafting the rules is intended to reduce the risk of incorrect responses or missed funds. The system also automatically creates comparison tables of old and new clauses from prospectus change records stored in the application, with bulk downloads in Word format.

The data platform, database, work interface, and calls to generative AI functions all operate within the existing Snowflake environment. Resona AM introduced a CI/CD workflow using GitHub Actions and Snowflake CLI to develop and deploy the system to production. In line with financial institutions’ operating principles, software handles tasks requiring certainty, while staff review and approve AI proposals. The work interface includes change highlighting, approve and discard buttons, and direct editing. The company says the system will also reduce the burden of creating and comparing old and new documents one file at a time.

Resona Asset Management announced on October 7, 2026, that it had built a system on Snowflake to automate the creation and revision of investment trust product summaries and prospectuses. Developed jointly with Nowcast, the system adds four functions to work previously handled by staff manually editing Word files for each fund: fund search, draft creation, bulk changes, and generation of comparison tables showing old and new text.

Changes to product summaries and prospectuses occur frequently when new funds are launched, internal rules are revised, or notices from groups such as the Investment Trusts Association, Japan, and changes in laws require updates. Responding to one revision notice previously took several days to about a week, and workloads grew with the number of funds. Storage of original Word files, naming conventions, and version management also depended on individual staff members.

The new system organizes a database of investment trust information and a folder structure in Resona AM’s existing Snowflake environment. It centrally manages information extracted from existing Word files as structured data. Users can filter fund searches by product category, investment region, and management style. They can also enter context in natural language, which AI converts into search criteria to identify similar funds for reference when creating documents or funds that need revisions.

A new-document and individual-change editor lets staff create product summaries or prospectuses based on similar funds, or revise documents for a specific fund. When staff enter change instructions, generative AI proposes draft wording that they can approve, discard, or edit directly on screen. The system can also export Word files with comments identifying the changed sections and their content.

For changes affecting multiple funds, AI drafts replacement rules pairing the original and revised wording based on revision notices or change requests. Software identifies the affected funds and scans their files; limiting AI to drafting the rules is intended to reduce the risk of incorrect responses or missed funds. The system also automatically creates comparison tables of old and new clauses from prospectus change records stored in the application, with bulk downloads in Word format.

The data platform, database, work interface, and calls to generative AI functions all operate within the existing Snowflake environment. Resona AM introduced a CI/CD workflow using GitHub Actions and Snowflake CLI to develop and deploy the system to production. In line with financial institutions’ operating principles, software handles tasks requiring certainty, while staff review and approve AI proposals. The work interface includes change highlighting, approve and discard buttons, and direct editing. The company says the system will also reduce the burden of creating and comparing old and new documents one file at a time.