
[StatGPT](https://statgpt.dialx.ai/) is an AI-driven Talk-To-Your-Data platform specializing in official statistical data. It is designed to help users and statistical organizations to query, transform, analyze, and interpret statistical data and publications using natural language interface. StatGPT is built on top of the DIAL platform, which enables enterprise-grade governance and security.

DIAL’s StatGPT application is an AI agent with a set of advanced tools, which, in a conversational manner, allow users to perform complex actions on large, structured datasets. By providing quick access to official statistical and economic data, analytical publications, and glossaries, StatGPT empowers users to make informed decisions and perform data analysis based on accurate and reliable information.

Agentic architecture of the StatGPT application allows it to respond to complex queries forwarding a request to a specific tool:

- **Data Query tool**: Builds and executes SDMX queries using the input from users in natural language.
- **Publications RAG tool**: Retrieves information from analytical publications (usually in PDF format) using the RAG approach.
- **Glossary tool**: Retrieves information from the organization’s glossary of terms and definitions.
- **Web Search tool**: Retrieves information from white-listed websites using the RAG approach.

To ensure transparency and determinism, users at each step can see the underlying reasoning of AI agents and links to the sources of information used to generate the response.

StatGPT can be integrated into a data portal, website or other applications to enable a seamless experience for users. Refer to [DIAL Documentation](https://docs.dialx.ai/video%20demos/Applications/dial-statgpt) to learn more.


Work with statistical and economic data in natural language using StatGPT – An AI-driven Talk-To-Your-Data platform

StatGPT


This case study describes the DIAL’s multi-agent AI assistant for [TimeBase](https://timebase.info/) – a high-performance time-series database created by [Deltix](https://www.deltixlab.com/).

Our assistant brings down barriers for working with market data by enabling even non-technical specialists to perform complex tasks and market data analysis using only natural language.

Users can input prompts in natural language to access market data from TimeBase, transform it, visualize the data on charts, and apply technical indicators for analysis. Our assistant seamlessly integrates with the communication platform of your choice, delivering real-time alerts about specific market events. At every step, users can see the AI agent's actions to validate its reasoning and ensure the logic behind its decisions meets the expectations.

Key highlights:

- **Data Retrieval**:  The agent generates Python code to fetch historical SPY prices (can be data for any other symbol stored in the database) from TimeBase and displays them on chart.

- **Technical Indicators**:  The agent calculates and adds technical indicators to the chart (in the video demo we used Bollinger Bands and RSI (Relative Strength Index but it can be any other indicator you need).

- **Real-Time Monitoring**: The agent launches a real-time TimeBase monitor in DIAL, applies the trading rules provided by the user in natural language, and sends alerts to the user through the chosen communication channel in real time.



Work with market data, calculate technical indicators and generate alerts in real time using AI-powered multi-agentic applications.

TimeBase AI Assistant


Financial advisors and analysts face a significant challenge in analyzing the massive volume of unstructured financial documents they receive. This manual process is time-consuming, error-prone, and can lead to delays to decision-making and potential oversight of critical risks. 

AI-powered assistants can transform this process by reducing the time spent on each document, minimizing the risk of human error, freeing financial analysts to focus on higher-value tasks, and ultimately enhancing the quality of insights delivered by financial advisors.

Key highlights:

- **Seamless Integration**: The AI chatbot powered by DIAL can be integrated directly into the primary application used by analysts, promoting rapid adoption and minimizing resistance to change.
- **Automated Document Pre-processing**: Leveraging Large Language Models (LLMs) trained on financial terminology, allows to automate the pre-processing of documents of different formats. This includes extracting key data points, identifying relevant sections, summarizing complex information, and providing initial risk assessments.
- **Accelerated Decision-Making**: By automating initial pre-processing tasks, analysts can dedicate more time to higher-level analysis, strategy development, and risk assessment. This helps to reduce decision-making time and enhance responsiveness to market dynamics. Analysts can interact with the AI assistant using natural language, generate assessments, delve into source documents, validate findings, and provide feedback to continuously improve the LLM's accuracy.
- **Enhanced Accuracy and Risk Mitigation**: LLMs can identify subtle patterns and anomalies often missed by human analysts, leading to more accurate risk assessments and potentially uncovering hidden opportunities.
- **Human-in-the-Loop Approach**: While automation is key, human oversight remains crucial. Analysts can validate the AI-generated output, provide feedback to the LLM, and ensure the final analysis aligns with their expertise and judgment.



Bring the efficiency of financial analysts to the next level with AI solutions that improve the way they work with documentation.

Economy & Finance

StatGPT

TimeBase AI Assistant

Streamlined Financial Analysis