improved

12.0 Release: AI Agents are here! Plus improvements to Document Intelligence, BPM and more

We are proud to announce our latest Major Update to the ElectroNeek Platform: Version 12.0.

This release marks a groundbreaking moment for ElectroNeek, redefining what's possible with Intelligent Automation with the launch of our AI Agents functionality.

What's New

AI Agents - A New Experience in Automation

We are thrilled to launch the first version of ElectroNeek's own AI Agents- custom-built, highly specialized, hyper-powerful process actors that you may incorporate into any of your processes to perform analytical, decision-making and content-generation tasks. The introduction of AI Agents allows for an entirely new class of use cases to be automated, such as those that did not easily translate into RPA Workflows due to their highly complex, human judgment-based logic. AI Agents can also interact with your existing and new RPA Workflows, leaving only the most mechanical, repetitive tasks for these workflows to perform.

AI Agents are included within the Processes & AI Agents module of the platform (previously known as Process Hub). Here’s what you need to know:

  • Zero Code Creation: Creating an ElectroNeek AI Agent is extremely simple -- you simply use your own words to describe the Agent's role / persona, tasks, goals, input and output formats, and any context that it needs to perform its function successfully.
  • Learn from Supporting Documents: After providing an AI Agent its instructions, you can upload documents with additional knowledge. These documents may be manuals, guidelines, FAQs, documentation, historical records, or any other type of data that enables the specialization and customization of your AI Agent. In version 12.0, only PDF, DOCX and DOC files are supported.

Creating a new AI Agent in the ElectroNeek User Portal

  • Put Your Agents to Work: After creating an AI Agent, you may include it in any of your Processes by adding the new "AI Agent" Process Stage. With this capability, your AI Agent can effectively communicate with Humans (via Forms), RPA Bots, and even other AI Agents, through the use of Process Variables.
  • Static and Dynamic Prompts: When setting up an AI Agent Stage, you must define the Prompt that the AI Agent will work with — and this prompt can include both static and dynamic text, including the ability to include Process Variables from any other part of your Process into the prompt, by simply selecting them from the Variables list, or typing their name in the "##variableName##" format. Process Variables will show up automatically in the list based on your other Stages and their corresponding variables.
  • Memory-Capable AI Agents: Within a Process, you may include as many AI Agent Stages as you want — and each new interaction may recall information from previous ones. By using Thread Variables, you can configure your Processes to include full-blown "conversations" between your process actors (Humans, RPA Bots, AI Agents) without the need to repeat or re-introduce any information. Threads are scope-limited to the Process Job in which they were started and cannot be carried over into other Jobs.
  • Additional Customizations: When setting up an AI Agent stage, you may also include additional "guardrails" — restrictions and limitations for the AI Agent's response to your prompt. These optional guardrails can complement any pre-existing limitations that are included in the AI Agent's own instructions. Additionally, you can customize the Process Variable that will hold the AI Agent's response, which you can later use to communicate the AI Agent's thoughts to humans (via Forms), RPA Bots (via Input variables), or other AI Agents (via their Prompts.)

Setting up an AI Agent Stage in your Processes


Document Intelligence - Smarter and easier to use than ever

Our Document Intelligence module (formerly known as Document Processing) has undergone a major upgrade, making it more powerful and user-friendly than ever. This module is the home of IntelliDocs, ElectroNeek's Intelligent Document Processing (IDP) engine.

  • Improved UI: We have cleaned up and improved the User Interface for all pages of the Document Intelligence module, improving User Experience and unifying the design language of this module. Additionally, we have improved the reporting of remaining page counts.
  • Bulk Web Recognition: Introducing a whole new way to leverage Document Intelligence, the Bulk Web Recognition page allows the uploading of document batches to be processed in the background by IntelliDocs, without the need to set up an RPA Workflow or an API call. These documents are uploaded into a Recognition Queue, and you can check their status in real time. As soon as each document is done processing, you may preview its results and download them in JSON format. When the entire Recognition Queue is completed, you may download all results for all documents together in a single package.

The new Bulk Web Recognition in action


Improvements to Process Management

The Processes & AI Agents module has also received updates to the pre-existing BPM functionalities.

  • Introducing Guest Users: You may now invite Guest Users into your Team. While regular Portal Users can be assigned permissions to use all modules of the User Portal, according to an organization's admin policies, Guest Users exist solely for use within the Processes & AI Agents module. Guest Users can be assigned the fulfilling of Forms within the Form Stages of your Processes without granting them control over the full process. This functionality makes it easier to include external users in your processes while maintaining security and oversight over the execution of your workflows.

Inviting Guest Users into your Team

  • Revamped In-Process Logic: The Logic Stage of the Processes has been expanded — instead of 3 pre-determined paths to evaluate a Process Variable's contents, you may now create as many customized Cases as you need. A Case is a specific condition in which you check the Process Variable against a specific Value, and if there is a match, your specified Action will occur. Actions include: Advancing to the next Stage, Jumping to any Stage in the Process, finalizing the Process Job with Success status, or finalizing the Process Job with Failed status. Aside from specific-value Cases, you may also add a Default Case — an Action that will be executed if none of the defined Cases match the selected variable's value.

The improved Logic Stage, with customizable Cases

  • Control Task Execution with Timeouts: You may now check the "Set timeout" option in any Process Stage to set up a maximum time in which the Stage needs to be fulfilled. You may customize how many days, hours and minutes will be waited for until the Process Job is marked as Failed due to a lack of response. If you do not check the "Set timeout" option, the Process Job will remain active until it is completed or cancelled.

Configuring Timeouts in your stages

  • Additional UI/UX enhancements: The Tasks and Jobs sections of the Processes & AI Agents module have received significant enhancements in the realm of UI/UX and usability. Both the Tasks and Jobs pages have received a faster refresh time, improving the flow of your management workflows. The Tasks page now includes both "Waiting" and "Completed" sections — the "Waiting" section includes Tasks that a user has been assigned to fulfill currently, while the "Completed" section includes all previous Tasks for that user. Additionally, Search and Filter functionality has been added to these pages.

Other Updates and Fixes

  • Revamped Billing Page: Owners and Billing Contacts on your Team can now see a visually improved and better-performing page with the platform's Usages and Limits.
  • New ElectroNeek Chrome Extension: We have updated the ElectroNeek Chrome Extension to comply with Google's rules regarding Manifest V3. The latest version of the extension is available in the Chrome Web Store. Please keep in mind that the new version of the extension is only compatible with Studio Pro and Bot Runner versions 11.4.5 and higher. Please make sure to update Studio Pro and Bot Runner accordingly!
  • Live AI Support: The Live AI Support feature, which we had incorporated into the Help Center, the User Portal and Studio Pro, has been fixed. This AI-powered Support chatbot will now answer your questions based on ElectroNeek Documentation and point you to its corresponding sources. As with most GenAI-powered solutions, it's prone to making mistakes if it doesn't understand your question or doesn't have enough information to answer it — so please let us know if you detect any inaccuracies in its responses!
  • Orchestrator Fixes: We have deployed fixes for a variety of reported errors that occurred when launching workflows from Orchestrator, such as launches getting stuck, disconnections, and unrequested duplicated launches of the same workflow.
  • Passing Files from Forms to RPA Bots: Fixed issues that could occur if a file of certain sizes was uploaded to a Form and passed as an Input variable to an RPA Bot within a Process.
  • Empty Process Stages: Fixed an issue where a Stage would be added to the Process even if no Stage Type was selected.A constant flow of new leads

As we continuously strive to adapt and evolve our technology to meet your needs, we’re eager to see how these new features can transform your day-to-day tasks. Stay tuned for more updates and continue sending us your feedback to help shape future enhancements. Thank you for being part of our community and for trusting us to help accelerate your digital transformation journey.