MyTrader is an AI research workspace that helps users analyse data, collaborate on code, organise sources and form reusable research reports.
From understanding the question and organising sources to analysing data, writing code and delivering a report, complex work needs a continuous path. MyTrader keeps every step connected in one workspace.
Move from data and code to a research result that keeps working.
Find changes, anomalies and relationships across tables, metrics and documents, then turn them into a clear summary.
Start from a requirement, existing code or an error and move through implementation, explanation, debugging and tests.
Break down a question, organise sources, compare viewpoints and preserve citations in an editable report.
Research is a structured progression, not a one-off answer.
Confirm the goal, scope and required result, turning a broad intention into a clear research topic.
Bring data, documents and sources together for an initial quality and usability check.
Explain metrics, collaborate on code, identify changes and open questions, and make the reasoning legible.
Deliver the summary, code, report and next action so each study leaves a reusable digital result.
Market research is one way MyTrader handles complex questions. The following example shows how a market-data question becomes a clear research report.
View the research processFollow the collaboration from the research question to the final memo. Select a step to focus the relevant part of the work.
| Field | Type | Sample | Status |
|---|---|---|---|
timestamp | DATETIME | 2026-03-15 | Normal |
close_index | FLOAT | 104.28 | Normal |
volatility_30d | FLOAT | 0.142 | Calibrated |
turnover_ratio | FLOAT | 0.038 | Normal |
sentiment_score | FLOAT | 62.4 | Normalised |
source_tag | STRING | IndexFeed | Verified |
import pandas as pd
import numpy as np
# 1. 整理时间窗口与对数变动率
df['log_ret'] = np.log(df['close_index'] / df['close_index'].shift(1))
# 2. 计算 7日 / 30日 滚动年化波动率
df['vol_7d'] = df['log_ret'].rolling(window=7).std() * np.sqrt(252)
df['vol_30d'] = df['log_ret'].rolling(window=30).std() * np.sqrt(252)
# 3. 按不同时间窗口聚合统计分布
summary = df.groupby(df['timestamp'].dt.to_period('M')).agg({
'vol_7d': ['mean', 'median', 'std'],
'vol_30d': ['mean', 'median', 'std']
})
print("波动特征聚合完成")Annualisation uses 252 trading days. The 7-day window reacts to short-term shocks; the 30-day window describes a more balanced cycle.
Core question: How do volatility characteristics change across time windows?
A multi-period decomposition of 14,280 continuous time-series samples shows clustered short-term volatility pulses, while windows above 30 days become steadier and show mean reversion.
Key findingsData comes from public time-series databases; cleaning and aggregation are fully defined in the analysis code.
Next question: does turnover quantile lead or lag the next volatility phase?
Move beyond one-off answers and build research assets that can be handed forward and verified.
Start in natural language, tighten the scope and define a precise research entry point.
Keep data, code and findings on one line so the reasoning can be replayed.
Continue editing and sharing the report, code and sources as the next study begins.
A specialist team and a research-first product help analysts move complex topics forward.
MyTrader keeps one question moving: find sources, analyse data, collaborate on code and finish with a clear report that can be edited again. It avoids unsupported predictions and black-box answers, freeing researchers from repetitive data preparation so they can focus on insight and reasoning.
Understand how MyTrader works and where it fits in a research process.
Yes. It can organise sources, check data, analyse metrics, collaborate on code and form a research report.
Yes. Start with a natural-language question and inspect the code and analysis when you need more depth.
Yes. The summary, code snippets, sources and report structure can all become the starting point for the next task.
Select “Start research” or contact the team to learn about the product entry point.