Document Processing
The biggest challenge was providing Japanese language support. Specifically, TIGEREYE's unique prompt rule design and LoRA fine-tuning were essential for accurately interpreting Japanese layout during optical character recognition (OCR) and managing the mix of honorific and colloquial language in Office Agent.
Reported impact
The following outcomes were attained by calibrating Qwen as a "local LLM" and leveraging high-performance processing capability for accelerated learning. TTS/Voice Cloning : Reduced learning time by 25%, significantly improving the naturalness of synthesized speech. Dual OCR : Improved form classification accuracy by 12%. This capability is currently being tested and implemented in local public service sectors and medical institutions.
TIGEREYE used Alibaba Cloud to address The biggest challenge was providing Japanese language support. Specifically, TIGEREYE's unique prompt rule design and LoRA fine-tuning were essential for accurately interpreting Japanese layout during optical character recognition (OCR) and managing the mix of honorific and colloquial language in Office Agent. TIGEREYE, Inc. is developing a multimodal AI platform designed to address social challenges leveraging AI's "eyes" and "ears", adhering to its vision of "from recognition to communication". The company provides AI agents that can "understand and communicate" in real time by integrating recognition-based AI, such as facial recognition, skeleton estimation, emotion analysis, speech understanding, and natural language processing (NLP), with large language models (LLMs) and speech synthesis technology. Alibaba Cloud was chosen for its processing power capability to run LLMs and a reliable support system.
TIGEREYE's workloads and source data are processed on Qwen, Model Studio.
Dual OCR : Improved form classification accuracy by 12%.
Specifically, TIGEREYE's unique prompt rule design and LoRA fine-tuning were essential for accurately interpreting Japanese layout during optical character recognition (OCR) and managing the mix of honorific and colloquial language in Office Agent.
TTS/Voice Cloning : Reduced learning time by 25%, significantly improving the naturalness of synthesized speech.
The biggest challenge was providing Japanese language support.
The following outcomes were attained by calibrating Qwen as a "local LLM" and leveraging high-performance processing capability for accelerated learning.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Alibaba Cloud Sample Jan 9, 2026 Bespin Global Bespin needed to industrialize enterprise AI agents without forcing customers to stitch together model stacks, orchestration frameworks, and multimodal data pipelines by hand. It also needed better document… delivery cycle reduced from 1 day to about 2 hours Alibaba Cloud Bailian · ChatOmni · CrewAI Alibaba Cloud Sample Archived · Oct 24, 2025 eSignGlobal High Compliance Costs: Navigating diverse legal standards across multiple jurisdictions increased operational expenses. Language Barriers: Complicated communication and interpretation of contracts in various languages… Automatically identifies legal risk points for over 70 countries and regions. Alibaba Cloud Elasticsearch · Anti-DDoS · ApsaraDB RDS Alibaba Cloud Sample Archived · Oct 19, 2025 Blockchain Solutions Limited Funverse addresses common e-commerce challenges by enhancing the shopping experience. For instance, consumers frequently add products to their carts but do not finish the transaction. Funverse leverages AI chatbots and… Elastic GPU Service · Elastic IP Address · Object Storage Service (OSS) 