01
Outline & task panel
Plan before executing, so you can see the research structure and the next action.
Beijing Starixs Technology Co., Ltd.
Deep Research,Stellar Insight
Beijing Starixs builds AI deep-research systems for business analysis, industry research, corporate strategy and consulting teams — research workflows that are traceable, verifiable, editable and deliverable.
Live product today
Raixis: an AI industry research assistant for professional research teams
Raixis is the AI industry research assistant from Beijing Starixs, built for industry research, company research, business analysis and strategy research. From a single topic, it quickly produces a well-structured professional report with traceable evidence and complete charts — one you can keep editing and export.
Unlike ordinary AI writing tools, Raixis doesn't just generate text or answer a question — it builds a complete workflow from research planning to report delivery around a professional task. You enter a topic; the system first produces an outline, and once you confirm it, automatically carries out retrieval, decomposition, data preparation, logical analysis, chart creation and content generation.
01
Outline first
Generate the research framework first; only after you confirm it does retrieval, analysis and writing begin — no one-shot black-box output.
02
Traceable evidence
Key steps are shown in real time, so you can question sources, add materials, adjust direction and review conclusions.
03
Editable delivery
Built for real business reports — keep refining structure and wording, then export deliverable results.
02
Transparent process & evidence
Tool calls, sources and analysis aren't hidden in a back-end black box.
03
Report editing & delivery
The final output fits real workflows — edit, polish and export.
AI doesn't replace researchers — it becomes an intelligent collaborator in research.
We integrate large-model capabilities, domain knowledge, data tooling and research methodology into one complete AI research system — cutting repetitive work, shortening cycles, improving report quality, and continuously accumulating reusable methods and knowledge assets.
Research planningSource retrievalData analysisReport delivery
Automatically organize goals, scope, chapter structure and key questions around the topic, so the AI output follows a professional research method from the start instead of producing loose text.
- Goals
- Decomposition
- Outline
Retrieval, citations, tool calls and interim conclusions stay visible, so you can question sources, add materials, adjust direction and review key judgments before delivery.
- Multi-source
- Transparency
- Source review
The system turns collected information into a comparable, discussable and verifiable analytical structure, combining data processing and charting for clearer judgments.
- Cleaning
- Comparison
- Charts
The final report is built for real business use — keep editing structure, refining wording and adding charts, then export, so results enter team collaboration and decisions.
- Editing
- Checks
- Export
- 01
Research planning
Start from a topic and break down goals, subjects, scope and report structure.
- 02
Source retrieval
Automatically gather multi-source materials around the question, with traceable entry points.
- 03
Information cleaning
Organize raw materials, filter noise and extract facts relevant to the analysis.
- 04
Logical breakdown
Separate facts, assumptions, variables and conclusions to make the reasoning path transparent.
- 05
Data analysis
Use data tooling for metric analysis, chart creation and quantitative reasoning.
- 06
Report validation
Complete writing, quality checks and wording refinement into a deliverable result.
Spanning deep research, knowledge capture, data engineering and business execution.
By spanning deep research, knowledge capture, data engineering and business execution, we use AI to turn scattered data assets into verifiable, collaborative and continuously evolving business capabilities — helping professionals and organizations make better decisions and achieve healthier operations and growth in complex environments.
- Flagship
Raixis
Professional research & industry insight
The flagship product for industry research, company research, business analysis and strategy research.
- Planned
Data engineering & quantitative reasoning · knowledge capture · business-process collaboration
Data foundation
Confident conclusions start with the data itself.
Raixis runs on a growing indicator library — data from exchanges, statistics bureaus, international organizations and industry providers, queried and charted on demand, with every figure cited.
~3M
indexed indicators
700K+
queryable indicators
~300M
data points
200+
authoritative sources
Covering 50+ industry and macro categories
Industries & commodities390K+
- Agriculture
- Building materials
- Steel
- Non-ferrous metals
- Energy & chemicals
- Textiles & apparel
- Light manufacturing
- Environment
- Commodity futures
- Transportation
- IT & telecom
- Utilities
Macro & statistics130K+
- National accounts
- Price indices
- Population & employment
- Business surveys
- Trade & balance of payments
- Fiscal & tax
- Money & credit
- FX
- Global economy
- Spatial & alternative data
Companies & markets170K+
- Companies & issuers
- Indices & quotes
- Equities
- Bonds & rates
Sourced from the Shenzhen and Shanghai stock exchanges, China's National Bureau of Statistics, the FAO, the World Bank, the OECD and more — 200+ publishers in all, queryable during research, with every figure cited.
Data architecture
Four layers of data, each one cited.
From our own indicator library to the files you upload, Raixis draws on each layer as needed — and every key figure in a report traces back to its source.
- 01
Indicator library
Structured indicators from exchanges, China's National Bureau of Statistics, the FAO, the World Bank, the OECD and more — queried, computed and charted on demand, with every number traceable to its publisher.
~3M indicators · 700K+ queryable - 02
Industry data partners
We work with Mysteel and other industry data providers across commodities, building materials, chemicals and energy, grain, healthcare, patents and commercial real estate.
11+ partners - 03
Live web retrieval
Targeted search over an allowlist of authoritative sites for policy, news and company updates, cross-checked across multiple search engines.
400 trusted sites - 04
Your own files
Upload PDF, Word, Excel or CSV files and analyze them alongside public data in the same report.
Multi-format upload
Every layer keeps its source, so each key figure in a report can be traced back.
Building the data foundation with leading industry data and intelligence providers.
Commodity prices, construction and engineering updates, chemical and energy quotes, coal market data, grain market data, pharmaceutical and healthcare data, patent and innovation data, commercial real estate and retail brand data, commodity indices and tender information — this high-value industry data forms the factual foundation of deep research. We work with the following organizations on data and market intelligence.
Why Raixis
Not just answers — research you can deliver.
General-purpose AI is great at quick answers. Industry research also needs a steerable direction, verifiable data and a deliverable result.
| General AI chat | Raixis | |
|---|---|---|
| Direction | One-shot output, hard to steer midway | Outline first — writing starts after you confirm |
| Data | Mostly model memory and general web pages | ~3M in-house indicators + targeted search over 400 trusted sites |
| Citations | Hard to verify; figures of unclear origin | Key conclusions cited one by one, one click to trace back |
| Process | Straight to the answer, steps hidden | Every search, computation and chart call can be expanded |
| Pre-delivery check | No dedicated verification step | Citations checked; unverifiable links and placeholders removed |
| Output | A block of chat text | An editable report, exported to deluxe Word / PDF in one click |
FAQ
Questions you may have
How is Raixis different from general AI chat tools?
General AI is built for quick answers. Raixis is built for complete industry research: it confirms an outline with you first, then queries its indicator library and trusted sites to pull data, build charts and write — citing key conclusions — and delivers an editable, exportable report.
Where does the data come from?
Four kinds: our indicator library of ~3M indicators, 700K+ of them directly queryable, from 200+ publishers such as stock exchanges, China's National Bureau of Statistics, the FAO, the World Bank and the OECD; partnerships with 11+ industry data providers including Mysteel; live retrieval across 400 trusted websites; and the files you upload. Every figure cited in a report shows its source.
Which industries can it research?
The library covers 50+ industry and macro categories — agriculture, building materials, steel, non-ferrous metals, energy and chemicals, textiles, environment and more, plus national accounts, prices, trade and employment statistics. Topics outside the library are supplemented through live web retrieval.
Can I trust the conclusions? How do I check them?
Key figures and conclusions carry their sources, so you can open and verify each one. Before delivery, citations are checked automatically and unverifiable links removed. Content is AI-generated — please re-check key data before important decisions.
Can I use my own materials?
Yes. Upload PDF, Word, Excel, CSV, Markdown or image files, and they are analyzed and cited alongside public data in the same report.
Can I export reports?
Yes. Reports can be previewed, edited and viewed as source, then exported to Word or PDF with cover, table of contents, headers, footers and formatted tables.
Bringing professional research toward human–AI collaboration, transparent process, traceable evidence and reusable results.
We believe professional research should not stay in the traditional mode of manual retrieval, organization and writing, but move toward intelligence, systematization and standardization. Beijing Starixs will keep deepening AI research workflows, knowledge engineering, data analysis and report generation.
Purpose
Use AI to raise research efficiency, professional logic to safeguard quality, and deep insight to create decision value.
Mission
Rebuild the deep-research process with AI, making high-quality industry research faster, more professional and more trustworthy.
Vision
Become a leading AI deep-research platform, delivering high-quality, intelligent and verifiable research solutions for industry research, corporate analysis, business decisions and strategy consulting.