<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>AI on Human in the Loop</title><link>http://aitip.tech/en/tags/ai/</link><description>Recent content in AI on Human in the Loop</description><generator>Hugo</generator><language>en-US</language><lastBuildDate>Tue, 17 Nov 2026 09:00:00 +0800</lastBuildDate><atom:link href="http://aitip.tech/en/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>AI Spec-Driven Development: Why We Chose OpenSpec</title><link>http://aitip.tech/en/posts/openspec-spec-driven/</link><pubDate>Tue, 17 Nov 2026 09:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/openspec-spec-driven/</guid><description>When you code with AI, requirements live only in chat history and vanish with the next session. Spec-driven development puts them in the repo, but Spec Kit, Kiro and Tessl take very different approaches, and critics call it waterfall coming back. This post compares the main tools against the industry critiques, explains why we chose OpenSpec, and shares the practices we built up over a dozen-plus repos and more than a hundred changes.</description></item><item><title>Character Consistency in AI Short Dramas: Asset Chains, Continuity Locks and Reference Images</title><link>http://aitip.tech/en/posts/ai-drama-consistency/</link><pubDate>Mon, 16 Nov 2026 13:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/ai-drama-consistency/</guid><description>When an AI animated series is made through APIs alone, characters, props and locations stay consistent across shots only through reference images and fixed text. Using real images from an original series, this post explains why source images are reviewed first, why model references are kept apart from human-facing sheets, why continuity locks are reused verbatim, why a location needs one plate per camera direction, and how to check shot by shot.</description></item><item><title>Designing an AI Short-Drama Pipeline: Shared Stages and Assets, One Skill per Genre</title><link>http://aitip.tech/en/posts/ai-drama-pipeline/</link><pubDate>Mon, 16 Nov 2026 12:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/ai-drama-pipeline/</guid><description>Paid short dramas, animated series, narrated novels and product ads each need a different rhythm and different checks. In this design, every genre shares the same four stages (script, assets, shots, video) and the same asset library, and what differs between genres goes into one skill that is layered onto every sub-agent. It compares open-source short-drama tools and multi-agent papers, with an interactive demo.</description></item><item><title>94% of Fields Correct. Now What? A Two-Stage Architecture for Scanned Documents</title><link>http://aitip.tech/en/posts/two-stage-document-extraction/</link><pubDate>Mon, 16 Nov 2026 08:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/two-stage-document-extraction/</guid><description>Across 38 scans and 4,732 field assertions, the classic OCR pipeline passed 94.08% of fields, but only six documents passed every check. Page transcription, extraction and large-schema experiments show how to separate reading from interpretation, retain source evidence, and avoid treating model agreement as proof of correctness.</description></item><item><title>Chunking for a Knowledge Base: One Markdown Convention, Chunked by Heading</title><link>http://aitip.tech/en/posts/pdf-to-markdown-chunking/</link><pubDate>Sun, 15 Nov 2026 11:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/pdf-to-markdown-chunking/</guid><description>What the knowledge base maintains is Markdown: PDFs, web pages, manual material and Q&amp;amp;A all follow one convention, and the chunker reads only the Markdown heading tree and block structure. This post covers the convention, PDF transcription and checks, the heading-level chunking rules, and the retrieval evaluation showing why one heading per chunk works best.</description></item><item><title>Building an Enterprise Knowledge Base: Content Model, Git Publishing and Hybrid Search</title><link>http://aitip.tech/en/posts/knowledge-base-architecture/</link><pubDate>Sun, 15 Nov 2026 10:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/knowledge-base-architecture/</guid><description>A knowledge base works like a library: categories, books and editions, tables of contents, passages, narrowing the context level by level to give the chat model accurate knowledge. Drawing on a bilingual project with about 1,800 documents, this post covers defining the content model first, using Git for review and publishing, keeping exact document fetch separate from hybrid search, and choosing retrieval methods with an evaluation set instead of intuition.</description></item><item><title>Designing an AI Agent for Filling Contracts</title><link>http://aitip.tech/en/posts/contract-fill-agent/</link><pubDate>Fri, 13 Nov 2026 16:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/contract-fill-agent/</guid><description>Built on official Chinese model contracts: AI pulls facts out of IDs, screenshots and chat with a source for each, code fills the fixed clauses, writes amounts in words and checks every rule, and people fill the gaps and settle conflicts before the contract is generated. With an interactive demo.</description></item><item><title>Designing a Template Reply Agent with Classification and Routing</title><link>http://aitip.tech/en/posts/template-reply-agent/</link><pubDate>Fri, 13 Nov 2026 14:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/template-reply-agent/</guid><description>Classify each incoming message, route it by priority to exactly one standard template, let the model only fill the blanks, check the draft in code, and have staff review and send. Shown with export sales inquiries, with an interactive demo.</description></item><item><title>UI Design for a Conversational Quoting Agent</title><link>http://aitip.tech/en/posts/conversational-quote-agent/</link><pubDate>Fri, 13 Nov 2026 12:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/conversational-quote-agent/</guid><description>Using an online EV quote as the example: the model understands and explains, pricing and finance services set the numbers, and cards carry every choice and confirmation. With an interactive demo.</description></item><item><title>AI as the Support Agent's Co-pilot: Designing a Customer Service Copilot Prototype</title><link>http://aitip.tech/en/posts/agent-assist-design/</link><pubDate>Fri, 13 Nov 2026 10:00:00 +0800</pubDate><guid>http://aitip.tech/en/posts/agent-assist-design/</guid><description>Instead of letting AI reply to customers directly, seat it next to the human agent — drafts, citations, clarification, guardrails and a feedback loop. Design notes on an insurance customer-service copilot prototype, with an interactive demo.</description></item></channel></rss>