Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
D.AT is a quantitative trading support tool launched by D.AT Analytics, LLC, positioned as “Data for AI Stock Price Predictions.” It does not claim to automatically generate stock prediction models for users. Instead, it standardizes some of the most time-consuming parts of stock machine-learning workflows—data engineering, label design, and backtesting/significance testing—helping users move into the modeling stage more quickly.
Based on the information on the site, D.AT covers data cleaning, time-series window slicing, multi-source data aggregation, feature engineering, strategy label generation, train/test splitting, and Backtesting/Significance. Its more valuable aspect is that it is designed around common issues in financial modeling. For example, when splitting data, it emphasizes reducing look-ahead bias and survivorship bias. Its labeling features can turn specific trading strategies into machine-learning targets, such as buying at the open and taking profit at 5% or stopping loss at 3% within 10 trading days. However, users still need to build the models themselves; D.AT is more of a data and evaluation infrastructure layer.
The page includes a “Try for Free” option, but it does not specify the free quota, trial period, feature limitations, or paid plans. The terms state that the platform may modify service fees at any time and notify users, so for now it is only possible to infer that there may be a free trial followed by a paid model. Long-term cost cannot be evaluated from the available information.
The main advantage is its focused use case. It offers a fairly complete workflow design around machine learning for stock prediction, especially for users who need to clean and aggregate multi-source data such as price, sentiment, and macro data, then build strategy-based labels. The downside is the lack of public information: there is no clear detail on API availability, integrations, data source coverage, exact pricing, Chinese-language support, or customer support channels. The terms also explicitly state that the platform does not guarantee the accuracy of external data, absolute data security, no data loss, or software backward compatibility. These are important risks to assess for production-grade quantitative systems.
D.AT is better suited to quantitative researchers, financial machine-learning developers, and trading strategy researchers with modeling capabilities. It is not ideal for ordinary investors looking for “one-click stock predictions.” The main text does not provide information on access from mainland China, and supported payment methods are also unknown. If access, compliance, or local data integration becomes an obstacle, alternatives include QuantConnect, Backtrader, Zipline, MLFinLab, or a self-built Python data engineering pipeline.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on d.at official site.
d.at is an Unknown Finance provider. TG4G tracks its product information, an overall rating of 7.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach d.at directly.