Turn information into usable knowledge
More information can make the useful part harder to find.
Connect content, context and retrieval across source reading, personal records and enterprise knowledge.
AI, MADE YOUR OWN.
AI capabilities, made your own.
Through accessible AI applications, digital assistants and intelligent workflows, we help people and organizations organize knowledge, connect their work and make room for creativity.
INDEPENDENT THINKING · PRACTICAL INNOVATION
TIAOTECH / LIUZHOU, CHINAFrom an individual idea to a shared way of working.
01 / TIAOTECH
Technology opens new possibilities.
Make intelligent work a capability people and organizations can shape.
Start with a specific need and bring technology into real work and life. Keep exploring useful, dependable software that turns ideas into products with lasting purpose.
Our mission and values ↗INDIVIDUAL CREATIVITYORGANIZATIONAL CAPABILITYLASTING PRODUCT VALUE
FOR PEOPLE
Less scattered. More considered.
Scattered records, a changing news cycle, everyday decisions. We start with specific needs to explore personal AI and mobile products.
FOR ORGANIZATIONS
Knowledge into action. AI into the business.
Connect knowledge, access, tools and workflows in real work.
02 / FROM PROBLEMS TO POSSIBILITIES
The next step for AI? Something useful.
Connect repeat work into intelligent flows
Switching between tools can disconnect actions from outcomes.
Connect AI, business actions and human confirmation around a defined task, with boundaries and records for each execution.
Build applications around the business
Each new requirement can repeat work on users, access and business foundations.
Use shared platform capabilities for administration, content, memberships and multi-client products, keeping reusable capabilities distinct from product domains.
03 / TIAOTECH PLATFORM
Business has many parts. Connect them.
An AI application platform for enterprises and development teams, bringing together access control, business operations, knowledge, agents, workflows and multi-client applications.
Implemented modules and local evidence · Environment-specific acceptance required
Explore the platformIdeas become real through experience.
Explore all products
01 / Mivori
Follow meaningful developments in AI and technology
An AI and technology reading product for Chinese- and English-speaking readers outside mainland China, being developed around source-linked articles, briefings and followed topics.
- THE NEED
- Fast-moving information is spread across sources, languages and topics.
- THE APPROACH
- Open the event and source stories to compare coverage.
02 / Yivora
Build everyday outfits from clothes you already own
A digital wardrobe product in development, connecting garment records, outfit planning and wear history so your wardrobe is easier to understand and use.
- THE NEED
- It can be hard to see what you own or keep favorite outfits and wear history together.
- THE APPROACH
- Use the occasion and temperature to adjust an outfit from your wardrobe.
03 / AI Ball Predictions & Tips
A clearer view before the match
Explore football and basketball information, AI-assisted analysis and reference tools to form your own view before a match.
- THE NEED
- Scattered information and isolated predictions can leave important context out.
- THE APPROACH
- Compare perspectives, open relevant tools and check inputs and entitlements.
Care for every step. From idea to use.
Understand
Identify the person, task and context. Turn an idea into a problem that can be discussed.
Prototype
Use interactive screens to examine flows, content and states before implementation.
Build
Connect interfaces, data and business capabilities, including access, failures and recovery.
Improve
Compare documentation with actual behavior and turn feedback into the next iteration.
06 / WHAT GUIDES US
Stay curious. Stay grounded. Keep moving.
Practical innovation
Start with a real problem. Choose technology for the practical value it can bring.
Build together
Learn through conversations, trials and feedback. Shape the solution with the people who will use it.
Useful and dependable
Make actions and outcomes understandable, with clear ways to recover when something fails.
Keep evolving
Build capability through validation and iteration. Turn each lesson into the next improvement.
Stay curious. Stay in the loop.
More news ↗AIHOT · Selected summaries with original sources · Oct 29, 2026, 9:00 AM
AI models · Latent Space
Elorian Releases Visual Reasoning Model Glimpse and Launches DecoyCount Benchmark ↗
Elorian released its first visual reasoning model, Glimpse, emphasizing that it participates in the reasoning process through image generation and editing; simultaneously launched the DecoyCount benchmark to test whether models truly count rather than relying on visual shortcuts; Glimpse scored 50.48% in an internal single-run test, being the only model exceeding 50%, while GPT-6 Astra, Gemini-3.8-flash, Qwen3.8-max, and Claude-Fable-5.1 scored 43.17%, 32.71%, 28.31%, and 25.22% respectively; all evaluation images and texts were not used for training Glimpse.
Practical tips · Cohere Blog
Cohere Releases Reference Architecture for Financial Research Assistant: A RAG Solution Based on Embed, Rerank, and Parse ↗
Cohere's official blog released a reference architecture for a financial research assistant, combining Cohere Embed (Pro/Fast), Rerank (Pro/Fast), Parse, and AWS services (SageMaker AI, OpenSearch Serverless, Bedrock). It supports page-level retrieval from PDF financial reports, table/chart understanding, multilingual queries, and answer generation with source attribution. The solution emphasizes verifiability (each answer is annotated with page number and content type), cost control (Embed Pro for indexing + Embed Fast for querying, int8 1024-dimensional vectors), and flexible toggle design (Rerank, Context Focus, and Page Image are all configurable options).
AI models · 月之暗面／Kimi
Moonshot AI Releases Kimi K3.1 Pro Large Language Model ↗
Moonshot AI released Kimi K3.1 Pro, the first open-weight large language model at the 3T scale, with a context window of 1 million tokens; simultaneously launched the accelerated version Kimi K3.1 Pro Highspeed, offering a 2–3x speed increase; the model achieved 65.7% on Terminal-Bench 4.0 and 61.2% on Kimi Code Bench 3.0 Long-horizon, improved financial modeling accuracy by 7.2 points, and reduced the hallucination rate from 23.4% to 8.3%; to control cyber offense and defense capabilities, relevant data was removed during training, causing the xdev-bench-hard score to drop from 90.5% to 81.0%; full model weights will be open-sourced on November 4, 2026.
Research · AMD ROCm Blog
AMD ROCm Blog Details Inference Optimization of World Action Models on AMD GPUs ↗
The official AMD blog introduces how to optimize the inference efficiency of World Action Models (WAMs) on AMD GPUs (MI355X/W7900) to support closed-loop robot control.
Technology products · InfoQ 中文站
Cloudflare Opens Fine-Grained Cache Control for HTTP Vary Response Headers to All Plan Customers ↗
Cloudflare has enhanced Cache Rules on its global edge network to natively support three handling methods for HTTP Vary response headers: normalization, pass-through, and cache bypass, helping operators avoid cache thrashing; this feature is now available to all plans including Free, Pro, Business, and Enterprise.
Perspectives · 虎嗅 · AI
Who is Dismantling US Chip Controls? ↗
On October 22, 2026, the US Treasury Department released a proposed rule to abolish the computing power threshold in AI investment reviews, upgrading restrictions on AI investment in China from tiered management to a principled prohibition. The article points out that this adjustment reflects three cracks in the original logic of computing power control: DeepSeek trained frontier models with low computing power, shattering the assumption that 'computing power = capability'; the scale of gray market smuggling is massive (for example, a $160 million H100/H200 network was dismantled); and companies such as ByteDance, Alibaba, and Tencent legally bypass controls by remotely accessing computing power through overseas data centers. Meanwhile, it analyzes NVIDIA's lobbying motivations and the reality of protected markets behind domestic chip price hikes and explosive performance, noting that controls are moving toward passive adjustments due to misaligned interests among multiple parties.
Start with what matters to you.
GET IN TOUCH
Get in touch ↗
- Company address
- 4-1, Building 1, Shengfu International, Liuyong Road, Liunan District, Liuzhou, China
- Operating company
- 柳州市跳动科技有限公司