450 emails received per day can now be handled simply by checking and sending. Automatic draft generation by AI dramatically frees up time to focus on specialized tasks.
Goodhill Systems Co., Ltd.
Issues before implementation
- Responding to emails required a lot of man-hours, putting pressure on other tasks.
- There was a risk of email replies being personalized and typos and omissions
- There was no system to raise productivity.
Effects after implementation
- Automating email replies significantly reduces response time
- Eliminating personalization leads to stable quality and fewer typos
- By incorporating AI into business processes, we can improve efficiency across the entire company.
table of contents
Company Introduction
Goodhill Systems Co., Ltd. is a digital archive company based in Tokyo that handles the digital conversion of analog video. They offer services to convert old videotapes and films into modern formats such as MP4 and DVD, and cater to both individuals and businesses.
We work steadily and carefully under the philosophy of "preserving our customers' important records in high quality." With a unique business model that combines video technology and IT, we have built trust and continue to grow by meeting the demand for digitization.
Issues before introducing generative AI
The biggest issue was the amount of time spent responding to customers. Dealing with phone calls, emails, and in-store visits occurred on a daily basis, making it difficult for the company to concentrate on specialized tasks such as video conversion. The number of emails handled, in particular, reached 450 per day, placing a heavy burden on staff and resulting in a decline in productivity.
Benefits of using generative AI
[Details of the installed system]
Digirise has created an "automated email support system," a tool to solve the problem of a large number of customer responses overwhelming specialized work.
The system works by linking n8n and Dify, crawling mailboxes periodically and having the AI automatically draft replies. For emails that require a response, the AI references a knowledge base containing the recipient's information and generates the optimal response.
This means that for simple inquiries and delivery date confirmations, which are particularly easy to create into templates, the person in charge only needs to check and send the draft created by the AI. The response time per message has been reduced to less than one minute, significantly freeing up time for the technical department to focus on their core business.
Future outlook
We plan to use the results of this email response as a stepping stone to "company-wide DX" and expand the scope of its use in accordance with the following two-stage roadmap.
Phase 1: Visualization of telephone responses andAnalysis of complaints
The most recent plan is to use AI to analyze the automated transcription data accumulated from daily phone calls, to detect signs of trouble early and share knowledge to improve response quality. This will reduce the burden on not only email but also phone responses while improving their quality.
Phase 2: AI-based image processing and digital transformation of the entire business
In the medium to long term, we are also considering introducing AI into core technical areas of our business, such as image processing (noise removal and editing assistance). While carefully assessing the technical and cost return on investment, we aim to incorporate generative AI into our company's overall systems rather than simply using it for partial optimization, leading to fundamental digital transformation.
