Document Processing · Enterprise Knowledge Search
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 understanding for Word, PDF, image, and Excel inputs and a faster way to import proprietary enterprise knowledge into production agents.
Reported impact
Bespin expanded supported corpora to more than 20 file types, increased image-parsing precision from 55% to 85%, improved overall accuracy by 50%, and reduced a typical multi-agent delivery cycle from about one day to roughly two hours, lifting overall service efficiency by 70%.
Bespin Global built its ChatOmni enterprise agent platform on Alibaba Cloud Bailian so it could package model access, evaluation, multimodal parsing, and RAG into one delivery stack. Bailian provides the model-training, deployment, knowledge-base, and moderation services, while Qwen-VL and Qwen text models handle image and document understanding for Word, PDF, image, and Excel inputs. Parsed content flows into a vectorized RAG layer that turns dispersed enterprise material into reusable knowledge, then ChatOmni wraps those services with customer-specific business logic and open-source orchestration using LangChain and CrewAI. Prompt-engineering utilities and review controls improve answer quality before responses reach end users. This lets Bespin ship document-factory workflows and enterprise knowledge assistants as configurable products instead of one-off integrations.
Word, PDF, image, and Excel files enter Bailian's OCR and parsing pipeline, where Qwen-VL and Qwen extract and normalize content. The processed material is vectorized into a RAG knowledge base, then ChatOmni uses LangChain and CrewAI orchestration, prompt templates, and review controls to generate answers, document outputs, or agent actions for enterprise users.
After embedding Bailian, ChatOmni expanded to more than 20 file types, raised image parsing from 55% to 85%, improved overall accuracy by 50%, and reduced delivery cycles from 1 day to about 2 hours. (original: 从传统的 pdf/word/excel 扩大到常见的 20 多种文件类型，并借助视觉模型识别将图片解析的精准度从传统方案的 55% 跃升到 85%... 整体准确率提升 50%... 交付周期由原本的 1 天压缩至 2 小时左右。)
Bailian's OCR and document-parsing engine, combined with Qwen-VL and Qwen, gives ChatOmni high-precision extraction and table reconstruction for Word, PDF, image, and Excel files. (original: 依托阿里云百炼集成的 OCR 与文档解析引擎，并融合 Qwen-VL 视觉模型与 Qwen 文本大模型，ChatOmni 实现了对 Word/PDF/图片/Excel 等多种格式文档的高精度内容提取与表格重建能力。)
Bespin built an enterprise knowledge hub on Bailian's knowledge-base and RAG engine by bulk-ingesting scattered materials and vectorizing them. (original: 基于阿里云百炼的知识库和 RAG（检索增强生成）引擎构建了企业知识中枢。通过批量接入分散资料完成向量化。)
Bespin said enterprises face high complexity choosing, combining, and maintaining the right large-model stack when they build AI agents themselves. (original: 企业在自建 AI 智能体时，面临如何从众多模型和开源框架中选择、组合和维护最优技术栈的挑战，这需要巨大的技术投入和试错成本。)
ChatOmni is Bespin's self-developed enterprise agent application platform for reconstructing office and business processes with AI. (original: ChatOmni 是由贝斯平自主研发的企业级智能体应用平台。该平台致力于运用 AI 技术对企业办公和业务流程进行重构。)
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.
The public company page's Organization JSON-LD includes numberOfEmployees value 336.
Related
Alibaba Cloud Sample Feb 23, 2022 Xiangnong Technology (香侬科技) Xiangnong Technology needed a cost-effective but production-grade AI stack for Chinese NLP products that span writing assistance, document intelligence, public-opinion analysis, and government workflows. As a… FireDragon Writing registrations grew 1500% in one year AKS · AMD instances · Cloud Monitor 
MongoDB Sample Sep 9, 2026 Squary AI Squary AI needed to eliminate slow, outage-prone legal search infrastructure that combined MongoDB with OpenSearch and hurt document review speed, cost, and trust. At Quatrini Law Group, Squary AI analyzed more than 460,000 pages and saved over 400 attorney and paralegal hours. Amazon Bedrock · Claude · MongoDB Atlas 
In-house Sample Sep 1, 2026 Thomson Reuters General-purpose LLMs can produce subtle but consequential errors in legal and compliance work, where professionals need trustworthy, citation-backed analysis across very large document sets. As of Q1 2026, CoCounsel has crossed one million users CoCounsel · Practical Law · Tabular Analysis 