Getting Started Guide

This document outlines the access requirements, deployment scope, interface structure, workspace configuration, dataset management, analysis execution processes, AI enrichment workflows, sharing capabilities, monitoring features, usage tracking, system limits, and troubleshooting procedures for the askEdgi platform across Public, SaaS, and On-Prem environments.

askEdgi enables natural language-based interaction with datasets, supporting metadata discovery, analytical execution, reusable workflows, and workspace-driven analytics without requiring coding.

askEdgi Platform Capabilities

askEdgi supports the following functional areas

• Natural language-driven analytics

• Workspace-based dataset organization

• Metadata discovery and profiling

• Analytical query execution

• Chat-based conversation tracking

• Recipe creation for reusable workflows

• AI-powered data enrichment

• Usage monitoring and workspace health tracking

Deployment Editions and Feature Scope

Supported Deployment Variants

Edition	Access Method	Core Capabilities	Limitations
Public Edition	Public askEdgi URL with self-service signup	Discovery and Analysis modes on public datasets and uploaded files. Analysis Mode is the default primary mode.	Subscription and plan limits
SaaS Data Analytics	OvalEdge environment with SSO or enterprise login	Full workspace analytics on catalog tables, connectors, and uploaded files	Tenant configuration dependent
SaaS Metadata Analytics	OvalEdge environment with SSO	Metadata analytics via OvalEdge-1 connector	No file uploads or raw data analysis
On-Prem	Hosted in internal infrastructure	Catalog search, profiling, metadata recipes, lineage	No file upload analysis or cloud AI enrichment

Access Requirements and Prerequisites

Public Edition Requirements

• Active account with a valid subscription

• Browser access to askEdgi

• Log in through the public askEdgi URL with a trial or personal plan

• Support for Discovery and Analysis modes on public and uploaded datasets. Analysis Mode is the default interaction mode.

SaaS Edition Requirements

• Access through the organizational OvalEdge tenant

• Authentication using SSO or enterprise credentials

• Full analysis in the Data Analytics tenants

• Metadata-only insights in Metadata Analytics tenants

On-Prem Requirements

• Hosted within internal infrastructure

• Limited to metadata analytics such as

• Catalog search

• Profiling

• Metadata recipes

• File-based analysis and cloud AI enrichment are unavailable

Initial Login and Landing Experience

Landing Screen Options

Option	Purpose	Mode
Watch a Demo	Embedded demonstration of workspace and analysis	Read only
Take a Tour	Interactive walkthrough of UI elements	Read only
Skip Guidance	Direct entry into the workspace	Interactive

AI Assistant Interface Structure

The primary interface contains

• Chat input panel

• Mode selector for Analysis and Discovery

• Response panel displaying summaries, tables, and charts

• Side panel listing My Chats, Recipes, Workspace objects, and sharing tools in Public Edition.

Interface Layout and Functional Areas

Area	Function	Description
Chats Panel	Conversation tracking	Stores prompts, responses, filters, and artifacts
Query Panel	Query execution	Displays text, tables, and charts
My Workspace Panel	Dataset management	Stores uploaded files and catalog datasets
Profile and Settings Bar	Account controls	Provides access to My Usage, Manage Subscription, and Workspace Status

Query Interpretation Modes

Discovery Mode

• Retrieves metadata such as tables, columns, glossary terms, governance information, and profiling summaries

• Supports structured catalog exploration and asset validation

• Does not execute data analysis

• Does not use Retrieval Augmented Generation

• Insights are metadata-based only

• Available across all editions, with scope dependent on deployment configuration

Analysis Mode

• Default and primary mode

• Uses Retrieval Augmented Generation to retrieve enterprise context before execution

• Validates dataset relationships and required attributes

• Executes analytical queries on workspace datasets

• Produces tables, summaries, and visualizations

• Available in Public and SaaS Data Analytics editions

• Not supported in Metadata Analytics and On-Prem editions

Asking Questions and Chat Context Handling

Mode-Based Example Queries

Mode	Example
Discovery	Find the sales orders table and show its description, owner, and data quality score
Analysis	Show total orders and average order value by month

Analysis Mode operates within a workspace-first execution boundary. Execution occurs only on pinned or eligible workspace datasets unless additional governed assets are required for validation.

Chat Context Behavior

• Each interaction runs inside a chat

• Chat stores

• Dataset selection

• Filters

• Pinned objects

• Previous responses

• Follow-up queries reuse existing context

• Context resets when a new chat starts

Important: Chat persistence enables progressive refinement without repeating dataset selection.

Adding Data to My Workspace

askEdgi requires datasets to exist in My Workspace before analysis. Data can be added through file upload or by selecting datasets from the Data Catalog, and it becomes immediately available for search, preview, and AI analysis.

Workspace Availability by Edition

• Public Edition supports uploaded files and public datasets

• SaaS Data Analytics supports uploaded files and catalog tables

• SaaS Metadata Analytics supports catalog metadata only

• On-Prem supports metadata only and does not allow file upload analysis

Uploading Files to Workspace

Supported Editions

• Public Edition

• SaaS Data Analytics

File Upload Steps

1. Navigate to My Workspace

2. Select Files to open the upload modal

3. Drag and drop or browse a local file

1. Wait for automatic validation

2. Confirm the file appears in the workspace object list

Upload Constraints

Constraint	Value
Supported Formats	CSV, PQT Parquet, XLSX, JSON
Maximum File Size	Up to 1 GB
Validation	Automatic format and size checks

Important: Unsupported formats or oversized files must be converted or reduced before upload.

Using Uploaded Files

• Files can be referenced in Analysis mode

• Queries can reference file names directly

Adding Catalog Objects to Workspace

Supported Editions

• SaaS Data Analytics

• SaaS Metadata Analytics

Steps

1. Open My Workspace

2. Click Catalog

1. Search or filter available objects

2. Browse using Explore Resources if required

3. Select objects using checkboxes

4. Click Add

Access and Validation Rules

Rule	Behavior
Minimum Permission	Data Read required
Unauthorized Objects	Hidden
Duplicate Objects	Blocked
Workspace Limits	Partial addition with warning

Live Query Execution

Live Query Execution allows askEdgi to run queries directly on supported source systems instead of ingesting data into the workspace. This improves performance and supports near real-time analysis.

Execution Behavior

• askEdgi selects the execution environment automatically based on data location and efficiency.

• Workspace Execution — runs inside the workspace when data is ingested.

• Source Execution — runs directly on the source system when live connections are configured.

Constraints

• Live and imported datasets cannot be analyzed together.

• Mixed execution is blocked until datasets are moved to the same environment.

• AI enrichment and transformations require workspace ingestion.

Organizing Workspace Content

Folder Management

1. Navigate to askEdgi > Workspace.

2. Click the three-dot menu beside Imported Objects.

3. Select Create a Folder.

1. Enter the desired folder name.

2. Click the tick mark icon to save and create the folder.

1. Folder names must

2. Be unique

3. Contain 1 to 256 characters

4. Use allowed characters only

Moving Objects

Steps to Move Objects

1. Select one or more objects in the workspace.

2. Drag and drop them into the desired folder.

3. Objects are automatically sorted alphabetically within the folder.

1. Move Objects modal for bulk moves

Removing All Objects from Workspace

1. Click Remove All from Workspace.

2. Confirm the action in the pop-up window.

3. All folders and objects will be removed from the workspace.

Pinning Behaviour

• When an object inside a folder is pinned, only that object is moved to the top of the folder.

• The entire folder does not move to the top of the workspace.

Note: Pinned objects move to the top within their folder only.

Initiate Analysis Focus

• Locate the pin icon labeled "Pin to focus your analysis" adjacent to each file or data catalog object within the workspace.

• Select the pin icon for each relevant item to focus analysis on that specific dataset.

Pinning Datasets for Focused Analysis

Pin Function

• Each dataset has a Pin icon

• Pinning ensures correct dataset selection

Pin Behavior

• Analysis restricted to pinned datasets

• Cross-dataset analysis is allowed only among pinned objects

• Unpinning restores automatic dataset selection

Important: Pinning is recommended when incorrect dataset selection occurs.

Chats and Conversation Management

Chat Storage

Chats preserve

• Prompts

• Responses

• Filters

• Dataset selection

• Analysis results

Chat Navigation

The interface displays a chronological list of all previous user interactions with the AI assistant.

• Access the conversation history panel to view past chats.

• Chats are organized with the most recent interaction at the top.

Chat Reactivation

Users can reopen any prior conversation Chat to continue their analysis from a previous point.

• Select a specific Chat from the history list.

• The chat interface reloads the complete conversation context, allowing for new follow-up questions.

Chat-to-Recipe Conversion

This function allows for the transformation of an existing conversation Chat into a reusable Recipe.

• Select the conversion option from the menu of a saved Chat.

• The system automatically preserves the analytical logic, queries, and configuration from the conversation.

• Open the Chats panel and select the desired conversation Chat.

• Click the 'Save as Recipe' option for the selected Chat to configure the Recipe.

• Share recipes with internal teams or publish to the Marketplace.

• Chats are displayed in reverse chronological order

• Selecting a chat reloads the prior context

• Context persists until a new chat starts

Important: Recipe execution follows deterministic behavior. Execution stops immediately when an error occurs and displays the exact error. Automatic retries or runtime corrections are not applied. If required inputs are missing, the system presents a controlled discovery step before execution.

Recipe Publishing

• Public Edition supports Marketplace publishing

• Enterprise SaaS supports internal sharing

Running an Initial Analysis

Standard Workflow

1. Add at least one dataset

2. Pin dataset

3. Analysis Mode is selected by default. Users may switch to Discovery Mode when metadata exploration is required.

4. Enter analysis query

5. Review generated tables and charts

6. Refine using follow-up questions

Sample Refinements

• Filter to the last six months

• Show only the top five product categories by revenue

Exporting Results

• Tabular results downloadable as CSV

• Export subject to format and size limits

Important: On-Prem and Metadata Analytics editions restrict insights to metadata only.

AI-Based Dataset Enrichment

Supported Editions

• Public Edition

• SaaS Data Analytics

Supported Enrichment Types

Category	Examples
Calculated Columns	Return rate, profit margin
Classification	High, Medium, Low segmentation
Text Analytics	Sentiment, intent, emotion

Example Enrichment Flow

• Create a return rate column using returns divided by orders multiplied by one hundred

• Classify products into High, Medium, and Low based on thresholds

Sharing Results Public Edition Only

Sharing Steps

1. Open an AskEdgi conversation Chat.

2. Click the Share Chat button located in the Chat header.

3. A pop-up modal will appear.

4. Select one of the following options:

5. Copy Link: Copies the shareable URL to the clipboard.

6. Post on LinkedIn: Opens a pre-filled LinkedIn post with the link.

7. Post on Reddit: Opens a pre-filled Reddit draft with the link.

1. Upon generating the link, a confirmation toast appears: “Link copied successfully.”

Sharing Behavior

• Shared chats display in read-only mode

• Charts and tables appear as static visuals

• Original chat remains editable

Monitoring Workspace Health

Workspace Status Indicators

Status	Meaning
Connected	Workspace active
Restarting	Reload in progress
Disconnected	Workspace unavailable

Workspace Control Actions

Action	Function
View Logs	Displays recent activity and errors
Reload or Restart	Restarts DuckDB session and clears cache
Upgrade	Increases memory, CPU, and table limits

Note: Uploaded files remain intact after restart.

Workspace and Subscription Limits

Workspace Object Limits

• Partial dataset addition when limits are reached

• Warning messages displayed

Plan and Spend Limits

Limit Type	Behavior
Public Plan	AI or compute tasks are paused when thresholds are reached
Enterprise Quotas	Organization-level usage tracking

Important: Corrective actions include removing objects, reducing workload size, or adjusting subscription plans.

Tracking Usage and Costs

Public Edition Usage

1. Click the Profile icon on the top navigation bar.

2. Select My Usage from the profile options.

My Usage

The My Usage feature in askEdgi provides a comprehensive summary of module consumption and associated costs. It enables organizations to monitor and analyze resource utilization across four key components:

• Recipes: Tracks the usage and associated costs of executed recipes.

• AI Tokens: Track AI processing usage.

• AI Enrichment: Monitor system enrichment usage.

• Compute: Infrastructure usage and associated compute costs.

1. Hover to view historical monthly usage

Subscription Management

1. Click the Profile icon on the top navigation bar.

2. Select Manage Subscription from the profile options.

1. Review Plans, Usage, Earnings, Billing

Enterprise Usage

1. Open Profile icon

2. Select My Usage

3. Review aggregated organization usage

Note: Detailed reports require administrator access.

Troubleshooting Common Issues

Issue	Cause	Resolution
Dataset not found	Missing dataset or no access	Add a dataset or request permission
Workspace limit reached	Maximum objects exceeded	Remove objects or upgrade
Empty results	Incorrect dataset or missing fields	Verify pinned dataset and column names
Slow workspace	Performance issues	Reload workspace or check logs
Recipe failure	Missing datasets or permission issues	Verify dataset availability, permissions, and inputs. Execution stops immediately when validation fails. Correct the issue and rerun.

Important: Execution logs should be reviewed before applying corrective actions.