A platform that allows non-technical users to ask data-related questions directly in Slack or Teams. It uses a data analysis language model to generate and verify SQL queries, helping data teams focus on essential analysis. The setup is quick and supports any database size and complexity, with all operations performed on-premises to ensure data security.
Features
Instant Answers for Non-Technical Users
Allows non-technical users to ask questions directly in Slack or Teams. The Data Analyst Language Model (DALM) then responds with the result, enabling users to get their questions answered immediately.
Generated SQL Query
Automatically generates SQL queries based on the questions asked. This helps in efficiently retrieving data from databases without manual SQL coding.
Supports Any Database
The platform is compatible with any database, size, or complexity. It enhances flexibility in data management and query execution.
Fully On-Prem & Local
Operates completely on-premise and locally, ensuring data privacy and security. Only schema and query code are sent to the LLM, not the actual data.
Installation
Detailed instructions for installing DaLMatian on Mac or Windows. It involves downloading the installer, running it, and opening the application.
Configuration
Instructions on configuring the workspace, including creating folders, copying necessary files, and connecting to a database by entering credentials or uploading a CSV. SQL queries are provided for specific database connections like BigQuery.
Workspace
Guidance on setting up workspaces focusing on customer-specific projects or contexts for query management and document organization.
Context Management
Allows for managing contexts by setting relevant queries or instructions that DALM should follow. Involves creating contexts with queries and rules.
Q&A
Features a Q&A section where users can enter a question, with the application retrieving relevant queries and explanations. Users can edit generated queries and save them.
Security
Ensures security by keeping the application on-prem/local, explaining that no actual data is accessed, only schemas, and clarifying OpenAI API usage without data retention.
Local Data Processing
The software processes input locally on your machine, without sending data externally, ensuring privacy. The only external connection is with OpenAI, similar to using ChatGPT.
File-based Query Training
Open a file with past queries to start training the DaLM, streamlining the setup process with no database connection needed.
On-Premises Security
Ensures data privacy by not tracking anything on the platform. Everything runs on your environment without external dependencies.
Data Privacy
No actual data leaves your environment. Only schema and query code are involved, ensuring data privacy and control.
OpenAI Key Integration
Option to use your own OpenAI key, promoting flexibility and security with zero data retention.
OpenAI Training Policy
OpenAI does not train on any data sent via API, ensuring data confidentiality from third-party interventions.
Workspace
Add files to your workspace like a traditional IDE. DaLM (Data Analyst Language Model) then indexes the content to inform SQL generation. Sharing feature enables DaLM to draw from your workspace to assist others on your team.
Ask Business Questions
Ask a question using language specific to your business. DaLM explains how context from your workspace is used to answer the question.
Generated Queries
View, edit, and run generated queries in the SQL editor. Save queries to your workspace with a click of a button – DaLM improves as more context is added.
Manage Context
View and edit the structured and unstructured context in your workspace. Provide additional business context for key queries.