
DataPeak Release Notes
Running List of New Features & Improvements
July 2025
July 18 - V.0.7.2
Custom Agent: PowerPoint Presentation Generator
Added a component for the Custom Agent that can create PowerPoint presentations through the use
of attached written documents and prompts provided by the user.
July 16 - V.0.7.1
Custom Agent: Forecast Suggestions
Added a new component that give forecast suggestions based on the accuracy of the comparisons.
June 2025
June 27 - V.0.7.0
Custom Agent: Allow for Custom Agent to create it's own components
Now allows for Custom Agents to create their own components for use on the system.
June 18 - V.0.6.5
AANA Forecasting Integration - DataPeak Implementation
Implemented a forecasting component into DataPeak that allows the user to generate a forecast
using machine learning models on historical data.AANA Forecasting Integration - Modal Improvements
Researched and prototyped modal improvements for the AANA forecasting experience to streamline
interaction and clarity.
May 2025
May 16 - V.0.6.4
Task Log Sidebar
Adds a page that lists a log of all actions taken in the system, including Custom Agent activity, dataset uploads, and more, so admins can quickly review what happened and when.
May 09 - V.0.6.3
Custom Agent: Iterator file compatibility
The Iterator now works with multiple input files. When a run produces multiple outputs, you can download one or all files, and export in different file types, making multi-file jobs simpler and more flexible.
Custom Agent Component: Allow for multiple input datasets
Lets the Input node in a Custom Agent accept multiple datasets at once, so an agent can work across several sources in a single run.
May 08 - V.0.6.2
Custom Agent Component: JSX Attribute Retrieval
Adds a component that retrieves attributes from JSX elements and returns them as output.Custom Agent Component: Group JSON
Adds a component that groups JSON data to help clean up and organize messy structures.
Custom Agent Component: Convert Nested Json to Structured JSON
Adds a component that converts deeply nested JSON into a more structured format for easier use.
Custom Agent Component: HTML input value parser
Adds a component that pulls values from HTML (e.g., inputs) so data can be extracted directly from HTML sources.
Custom Agent: Find relationship in data and group together using relationship
Adds a component that detects relationships in your data and groups items based on those connections.
May 05 - V.0.6.1
Custom Agent Component: Data Relationship Comparison
Adds a component that compares two related datasets/fields to identify how records connect, highlight differences, and help merge related items when appropriate.
Create Report AI Agent
Adds a Report AI Agent that can generate reports from selected data and export as PDF or save/share the report inside the system for other users.Research and integrate IBM Cloud Object Storage
Integrates IBM Cloud Object Storage as a supported data source, so teams can connect buckets and use them as dataset inputs inside DataPeak.Create AI Workflow Automation Agent
Creates an AI Workflow Automation agent that can run multi-step tasks automatically, including tasks like processing data and generating graphs as part of the workflow.Create dataset upload for SAP
Adds a dataset upload button for SAP, letting users authenticate and connect their SAP data directly into DataPeak.
May 04 - V.0.6.0
Speech to Text item for Custom Agent chat
Adds speech-to-text input to both the Custom Agent generation chat and the regular chat, so you can speak your prompts instead of typing.Custom Agent: Stream creation of Custom agent using chat
Enables live, step-by-step streaming of a Custom Agent being created from chat prompts, so you can watch components and settings appear in real time as they’re added.
Data Merge: Allow for adding AI Suggestions to Column Selection
In AI Data Merge, once suggestions are generated for which columns to join, you can now apply those suggestions directly to the Column Selection step, saving time and reducing manual picking.
Create starting templates for getting started with agents on platform
Adds starter templates built from existing user-created agents. You can pick a template and customize it, making it much faster to get up and running with agents anywhere in the system.Custom Agent Component: CSV file cleaner
Adds a component that parses CSVs more reliably by separating values on commas while ignoring commas inside data fields, reducing broken columns and cleanup work.Custom Agent Component: Data Merger for Inventory
Adds a Data Merger component for inventory workflows so outputs from multiple nodes can be combined into one, solving the prior limitation where you couldn’t merge results across nodes.Custom Agent: Debug Process
Adds a debug run for Custom Agents: you can execute steps and see how they behave before saving configurations and running the full agent, making it easier to spot issues early.Custom Agent Component: Expiration Date Tracker
Adds a component that tracks expiration dates in your datasets, especially for perishable inventory, and helps prioritize items that should be used before they expire.Custom Agent Component: Stock Optimization Component
Adds a component for stock optimization within Custom Agent workflows, helping set reorder points and safety stock levels so inventory stays balanced without stockouts or excess.Custom Agent Component: Demand Forecasting
Adds a component that lets Custom Agents predict future demand for a chosen period using historical sales data, so you can plan inventory and operations with more confidence.Make custom AI agents using chat
Lets you build Custom Agents directly from chat: describe what you want in natural language and the system assembles the agent for you. You can still drag-and-drop to edit if you want finer control.Custom Agents: Integrate Agent Deployment Notifications
Lets you add deployment notifications to a Custom Agent, so when the agent runs it can notify the agent owner, specific emails, or linked user accounts automatically.Custom Agent Component: Allow emailing results
Adds an option in Custom Agent workflows to email the results of a run, including setting the content, subject, and recipients right inside the workflow.Custom Agents Component: Allow for generating charts/graphs using data
Adds an option in Custom Agent workflows to generate charts and graphs directly from the data flowing through the agent, so visual outputs are produced as part of the run.Custom Agent Component: Allow agents to give text summary of data
Adds an option in the Custom Agent workflow to output a plain-text summary of the data the agent receives, handy for quick overviews before deeper analysis or exports.Custom Agents Component: Add option for creating reports
Adds a reporting option to Custom Agent workflows so an agent can export results as PDFs, Word docs, or Excel files directly from the run.Allow for editing, deleting, and publishing of agents
Lets users edit, delete, and publish Custom Agents so they can be shared and used by others across the workspace.Dataset integration Options
Researches and sets up data connection options so customers can choose their storage and pull data into DataPeak from the most appropriate sources.
April 2025
April 30 - V.0.5.10
Allow for using voice to text in Agent creator and Chat
Adds an option to use voice-to-text in both the Custom Agent creator and the main Chat, so users can speak prompts instead of typing.
April 22 - V.0.5.9
Implement Microsoft Office data connectors
Adds native Microsoft Office data connectors, enabling DataPeak to connect to and pull data from Office sources for use in workflows and analysis.Workflow Automation Options
Explores options for configuring automations in DataPeak, defining actions and triggers so workflows can run automatically based on preset conditions.
April 21 - V.0.5.8
Custom Agent: Add component to search internet for information
Adds a component that lets a Custom Agent search the web and return relevant information to answer a user’s question.
April 14 - V.0.5.7
Research swapping AI client to Cohere, Ollama or other option
Explores replacing the current OpenAI setup with Cohere, Ollama, or another provider, evaluating performance, cost, and a potential Canada-based stack to reduce reliance on U.S. systems.
April 10 - V.0.5.6
Custom Agents: Allow the agent to run either on a schedule or a specified number of times
Lets you choose how often an agent runs: set a recurring schedule (e.g., hourly, daily) or run it a fixed number of times, giving you control over timing and recurrence.
April 07 - V.0.5.5
Allow for pushing data to Airbyte Connector destinations
Enables pushing data out of DataPeak to supported Airbyte destinations (per the registered connectors), so pipelines can write back to external systems, not just read from them.
March 2025
March 13 - V.0.5.4
Custom Agents: Allow for the looping of sections
Introduces an Iterator component so an agent can loop a section of its workflow, processing multiple rows or multiple input files one-by-one through the same connected steps.
March 06 - V.0.5.3
Custom Agent Component: PDF Download
Adds a component that lets a Custom Agent generate and download a PDF of its results, including summaries, charts/graphs, and other outputs, so you can share or archive the outcome of a run.
March 03 - V.0.5.2
Agent Execution Log page
Adds a dedicated page that lists logs for every agent run, so you can review past executions and diagnose issues without digging through system logs.
February 2025
February 10 - V.0.5.1
Custom Agent: Add searching for components
Adds a search bar to the Custom Agent creation flow so you can quickly find specific components across all categories.Custom Agents Component: Integrate LIDA for asking questions of agents and generating infographics
Adds an option to use LIDA with Custom Agents for asking questions about data and generating infographics/visuals, similar to Auto-Analytics.Setup new User Management Hierarchy
Revamps the User Management Hierarchy so admins can mirror their company structure and assign role-based permissions to individual users more clearly.Allow chat to give summary of agents
Adds an option in Chat to summarize your Custom Agents, showing what each agent does, plus quick links to the agent itself and to the full agents list.Setup Data Management sidebar for uploading and modifying datasets
Adds a Data Management sidebar to streamline uploading and editing datasets from one place.Data Preprocessing Agent Features
Adds new preprocessing capabilities to the agent: identify column types, show sample values, count unique values, and compute min/max/sum, making data inspection much faster.
January 2025
January 29 - V.0.5.0
Custom Agents: Integrate Data Connectors
Allows Airbyte data connectors to be used inside Custom Agents, so agents can pull from connected sources directly.
November 2024
November 06 - V.0.4.0
Create resource allocation agent
Creates a resource allocation agent that assigns resources based on historical data, helping plan workloads more effectively.Data Cleaning and Preprocessing Agent
Creates an agent that cleans messy data into the right columns and rows, standardizing structure before analysis or export.API dataset improvements
Improves API datasets with options like setting keys for the response path and selecting which array values to use, making extraction more precise.Allow DataPeak to generate output API
Allows DataPeak to generate output APIs from entered data, so results can be accessed programmatically.
June 2024
June 21 - V.0.3.1
Allow for Admins to set fixed targets for all graphs
Lets admins define fixed target values for graphs, based on the grouping used in each chart.Allow for setting the min/max range for graphs
Lets admins define minimum and maximum value ranges on graphs during creation, so axes reflect the desired scale.Allow for duplicating data sheet templates
Lets admins duplicate existing data sheet templates so new versions can be created quickly without rebuilding from scratch.
May 2024
May 8 - V.0.3.0
Allow for chart/graph/table creation and mathematical operations to be performed with simple commands
Lets chat create charts, tables, and run math from simple commands, and improves guidance when something can’t be executed, so users get clearer feedback on what went wrong and how to fix it.