1. Modular text analytics

• Keyword and keyphrase detection
• Sentiment analysis
• Company name detection
• Brand and product name detection
• Custom machine learning classifiers
• Discovery
• Information extraction/parsing

2. Dashboard

Big data is not just about algorithms and processing. It is critical to have a good UI to understand how results are derived from text content, to edit data, and to provide feedback to improve performance.

3. Customize

Create and edit dictionaries, select processing modules, and tune performance. Custom algorithms can be developed and deployed within the Atomic service infrastructure.

4. Multiple content types

Our system works with multiple types of unstructured content including web-based content, specialized document collections, regulatory filings, and third-party content subscription services such as financial news.

5. API

Rich API for upload of content, access to analytics data, and processing configuration to support integration with clients' systems.

6. Backtest

Re-run analytics on batched content. For financial applications, analytics can be used to backtest trading strategies with linked financial market data.