Short questions.
Deep analysis.
From revenue mix to peer comparisons and multi-year growth. One question connects scattered filings and financials.
RESEARCH, IN CONVERSATION
Why it rose, why it fell, and whether anything in the financials needs a closer look. Ask across filings, financial intelligence and the newswire, and compare published figures on a chart.
This example includes English translations of the stored research. Where a translation is unavailable, a link identifies the original language.
Start with one of these
Stock in this conversation
PANW · Recently collected 8-K
See all data for this stockThis demo builds answers from existing public analyses. It does not generate new AI analysis, and it withholds judgment on questions outside the available data. As of 2026-10-21
See a tailored research demoChoose a stock to load its existing analysis.
THE VERTICAL AI FOR INVESTMENT RESEARCH
PHOENIX / 01Research that connects filings, macro, newswires and financial statements.
It produces target prices and reach probabilities by horizon,
and puts the data behind each call right on the chart.
005930 / KOSPI
Hover the chart to see prices by date. With a keyboard, use the arrow keys.
Past price history is shown with the published model target. 75% is a model estimate; it does not mean a validated hit rate, nor a guarantee of returns or of reaching the target.
What price, over what period, with what likelihood. Read the result of the analysis at once.
Markets
AI that speaks the language of markets.
It starts with data you can trust.
BUILT FOR THE WAY YOU RESEARCH
Cut the repetitive collecting and sorting,
and focus on the questions that need your expertise.
From revenue mix to peer comparisons and multi-year growth. One question connects scattered filings and financials.
From the answer back to the original DART or SEC filing. Ratios are computed in code, with the source and as-of date for every figure.
Target price, horizon and the probability of reaching it within that horizon. See the output of PheonixAI’s own model on the chart.
GROWTH & PROFITABILITY
Quarterly revenue and operating income, year-over-year change and operating margin.
The chart below is drawn straight from a Data API response.
3-year revenue CAGR+29.6%FY2022 → FY2025
Latest quarter26Q2
Tap a bar or use the arrow keys to move between quarters.
| Example companies | 24Q3 | 24Q4 | 25Q1 | 25Q2 | 25Q3 | 25Q4 | 26Q1 | 26Q2 |
|---|---|---|---|---|---|---|---|---|
| Revenue | ₩17.6T | ₩19.8T | ₩17.6T | ₩22.2T | ₩24.4T | ₩32.8T | ₩52.6T | ₩79.3T |
| YoY | +93.8% | +74.8% | +41.9% | +35.4% | +39.1% | +66.1% | +198.1% | +256.8% |
| Operating income | ₩7.0T | ₩8.1T | ₩7.4T | ₩9.2T | ₩11.4T | ₩19.2T | ₩37.6T | ₩60.5T |
| YoY | Turned profitable | +2,235.8% | +157.8% | +68.5% | +61.9% | +137.2% | +405.5% | +557.2% |
| Operating margin | 40.0% | 40.9% | 42.2% | 41.4% | 46.6% | 58.4% | 71.5% | 76.3% |
Source: DART periodic reports (quarterly, half-year and annual) · checked 2026-10-21
Operating margin is operating income ÷ revenue for the same quarter. A fixed example saved from a response received on 2026-10-21. Not investment advice.
WHY PHOENIX AI
Six things teams check when they choose a research data source.
Today's filings first
New filings are collected as they are accepted and turned into summaries and figures. Backfill work is scheduled so it never pushes today's filings back.
Same-day DART · SEC filings first
Keys on your server, managed in the console
Create API keys in the developer console and rotate or revoke them at any time. Accounts are issued after review, and remaining quota comes back in response headers.
Bearer auth · key rotation · quota headers
A source and a formula for every number
Each value says whether it came from DART, the SEC or company IR materials. Growth rates come with their base period and formula; when the base is a loss, you get “turned profitable” or “turned to loss” instead of a percentage.
source · formula · base period
From filings to earnings calls
Financial statements, growth and ratios, revenue mix, dividends, filing summaries and insider trades, all in one format. Earnings calls are in the PRO research workspace.
Korea · US available / Japan · China coming soon
Connect with an OpenAPI spec
Start with a single REST request. Generate a client from the published OpenAPI spec, or add it as a tool in Claude, ChatGPT or Cursor over MCP.
curl · fetch · requests · MCP
Built around your team's workflow
We design briefs around your holdings and mandate, and the screens and integrations your team uses. Start with a PoC on your own watchlist: 7 days by default (the term can be adjusted by agreement).
Enterprise · PoC, 7 days by default
THE RESEARCH BEHIND THE RESULT
Original filings, financial figures and company comparisons.
Review the source and basis of each analysis, then ask the next question.
FY2025 revenue was $416.2B. iPhone makes up 50.4% of total revenue, and Services forms the second pillar at 26.2%.[1]
| Segment | Revenue | Share | Source |
|---|---|---|---|
| iPhone | 209,586 | 50.4% | |
| Services | 109,158 | 26.2% | |
| Wearables, Home & Accessories | 35,686 | 8.6% | |
| Mac | 33,708 | 8.1% | |
| iPad | 28,023 | 6.7% |
Filing-based analysis example · Use the tabs to see revenue by product and by region, and the source document.
FROM COMPANY TO PORTFOLIO
Holdings, mandate rules, new filings.
Start the day with a morning brief built on your team’s context.
Held 33.0% / Mandate cap 30.0%
Review earnings assumptions against filings
Prioritized by portfolio exposure
EARNINGS CALLS
We transcribe the call recording a company publishes, timestamp by timestamp, and pull out AI key points.
Every number in a key point is checked against the passage it cites.
US FILINGS NEWSROOM
A PRO feature for financial newswires and newsrooms.
It turns SEC filings, such as earnings 8-Ks, into Korean-language article drafts. Your editors decide what gets published.
Available to invited or contracted accounts. During a PoC you can read drafts that have already been made, but you cannot create new ones. Pricing is by agreement.
The flash edition is a headline and two sentences. The standard edition has key points, an earnings table and the body. The in-depth edition adds a longer quarterly trend and, when available, earnings call highlights.
Tap a number or quote in the draft to open its source. Table cells show their row, column and unit declaration; quotes show the original English next to the translation.
Download the edition you are viewing as a file and move it into your editorial system.
The source line and the “AI-written draft” notice are always included, on screen and in every downloaded file.
FOR DEVELOPERS
Filings, financials, quotes and calendar data, plus PRO conversational research,
over REST and MCP. Generate clients from the OpenAPI specs.
PRICING
Start with the scope you need.
Want to see the workflow first? Book a 30-minute demo
Connect filings, financials, quotes, news and calendar data to your systems.
We design the screens and workflow around your team’s research.
For personal research programs and AI tools over MCP.
Data API, Enterprise and PRO API are all priced on request. Email help@alpsoft.io to get started. A PoC runs for 7 days by default (the term can be adjusted by agreement), and accounts are issued after review. Coverage is Korea, the US or both; Japan and China data are coming soon.
Frequently asked
Answers about research scope, evidence and bringing Phoenix AI to your team.
Have another question?Phoenix AI is a securities AI that researches and compares companies using Korean and US filings, financials and market information. It supports the repetitive research work of finance professionals, from gathering data to checking the evidence.
It is built for teams that repeatedly look up financials, compare companies and review filings: analysts, research teams, advisory firms, asset managers and IR teams. It cuts the time spent organizing data so the team can focus on the issues.
Yes. Korean companies are covered from DART filings and US companies from SEC EDGAR filings. You can compare companies across both markets and check the financial figures and filing evidence side by side.
From natural-language questions, it looks up a company’s financials and compares results and ratios across companies. It organizes multi-year trends and growth rates and presents the formulas and original filings so you can keep reviewing.
Yes. Filing-based answers include their sources, such as the report name, filing date and a link to the original. For financial ratios, you can review the figures and formulas used in the calculation.
It is reference content that gathers a stock’s price moves, key issues, market reaction and related articles in one place. Use it to scan scattered information and find issues worth checking further.
It is the target-reach probability that PheonixAI calculates with its own model. The target price, horizon and probability are provided together, and Phoenix AI receives this model output and displays it on the chart. For example, 75% within 1 year is a model estimate of the likelihood of reaching the target price within that period. It does not mean a validated hit rate, nor a guarantee of returns or of reaching the target. The chart on this page is a fixed example built from actual price history as of the date shown and published model output; changing the price-history range does not change the model’s horizon or probability.
Search for a stock and it loads the public analyses stored in Phoenix AI. You can explore approved analyses of recently collected 8-Ks, Korean financial intelligence, filing-based risk checks, published price-move analyses and the newswire through questions. “Show it as a chart” displays the available results comparisons or financial metrics as a bar chart and a table, with units and basis. The background to a rise or fall is shown with the news coverage of the time and the analysis date. It is a way to explore existing analyses: it does not generate new AI analysis, and it withholds judgment when data is missing.
Request an account at openapi.chickstockfi.com/en/request. We review each request and issue an account; there is no self sign-up. Log in to the developer console with that account and create your own API keys. A PoC gives you the Enterprise endpoints for 7 days by default (the term can be adjusted by agreement). Data API, Enterprise and PRO API are all priced on request. The quickstart at openapi.chickstockfi.com/en/docs/quickstart walks you from key to first request.
Book a demo meeting or send us an inquiry. Tell us how your team researches, the markets you follow and the work you need covered, and we will scope the rollout and timeline with you.
No. All data and summaries are provided for reference only and are not investment advice or a solicitation to buy or sell. Investment decisions and their consequences rest with the user.
LET’S GET TO WORK
Try Phoenix with your team’s watchlist and questions.
From a product demo to a PoC (7 days by default), we design it with you.