Hello Sumo-lings! 👋
Today, we are excited to share updates on our journey at Robomotion and what's ahead. Our story began five years ago in a pre-AI world and has since evolved into a powerful automation tool combined with cutting-edge AI capabilities.
A huge thank you to all our amazing Sumo-lings in our Discord community over the years. Your invaluable feedback has shaped our product into what it is today, driving us to innovate and evolve every step of the way.
─────────────────────────────
WHAT'S HERE NOW
─────────────────────────────
✅ Agentic Assistant:
Our Assistant has undergone a complete transformation. It now supports both interactive and agentic modes, offering an intuitive, chat-driven experience that simplifies complex tasks. The agentic mode provides a ChatGPT-like interface, making it effortless to create, manage, and interact with your automations. The interactive mode remains similar to the previous experience, guiding users with selections, inputs, and text. A new feature is our Custom Widget support, which lets you create and add custom components within the interactive chat. Instead of asking ten separate questions, your custom widget can collect all the necessary information in a single, stylized form and submit the data directly into your automation. This Custom Widget support will be further enhanced with generative, prompt-based AI in our upcoming new Designer.
✅ Agent Package:
Built on Google's Agent Development Kit (ADK), our Agent Package empowers you to integrate custom tools and create smart digital assistants — Agents — that combine the intelligence of AI with the reliability of RPA. This means your agents can perform multi-step tasks reliably by leveraging AI to enhance RPA's deterministic foundation.
✅ MCP Package:
Our new MCP support transforms your flows into MCP Servers, enabling seamless connections with various MCP hosts and clients like Cursor, Windsurf, or Claude Desktop. This not only streamlines processes but also opens new avenues for integrating your automation flows across different platforms.
✅ Agents Dashboard:
We’ve replaced our old Apps dashboard with a unified Agents Dashboard. This streamlined interface makes it easier than ever to manage your agents and automation flows in one place, delivering an improved and efficient user experience. It provides quick access to all your agents and serves as a gateway to our store, where you can install shared agents with one click.
✅ Agent Store:
The previous Apps concept was originally designed to serve as the backend for your custom applications, but the process was challenging for both non-technical and technical users. You had to use FlutterFlow to develop these apps, which was time-consuming and limited us to producing only three apps. Transitioning from Apps to Agents will change that. You will be able to create Agents for different tasks, similar to separate apps, but developed easily using the drag-and-drop interface of the Designer, eliminating the need for external development tools like FlutterFlow.
With the new Assistant, every Agent will have the same intuitive interface and experience. Additionally, with our Generative UI support, the overall experience will be even richer. GenUI enhances the traditional text-only chat interface with custom widgets for inputs or data display.
We will continuously add more Agents to our Agent Store and offer them for free, complete with customizable flows. The success of the Outreach Video Creator app, used and modified by cold outreach agencies to suit their needs, demonstrates the value of sharing these complex automations in our store. They were not just templates but fully functional apps. With the introduction of new Agents, this process will become even easier, allowing us to develop them more quickly. Our plan is eventually to open the Agent Store to everyone, enabling users to publish their Agents, whether for sale or as one-click installations for their clients.
✅ New Project Format:
To fully embrace the GenAI prompting world, we’ve reengineered our project format from a complex JSON blob to a text-based flow format. This change was essential for integrating AI to generate flows with prompts. It took significant effort to maintain backward compatibility, but we have now released our new project format along with our Git support. The upcoming Designer will let you create flows using prompts, offering a Cursor-like experience for your RPA automation development.
✅ Git Support and Better Collaboration:
A technical detail: we now have a Git backend for our flows. You don’t need to set up or use an external Git server like GitHub; every save is now a Git commit in our own backend. This means every flow is also a Git repository, allowing us to offer enhanced collaboration features, including detailed history tracking. Today, you can see who developed each part and when, and you can even revert to a previous state of your flow in time.
─────────────────────────────
WHAT'S COMING UP
─────────────────────────────
🚀 New Designer:
Prepare for a modern, collaborative canvas experience. Our new Designer will offer true real-time collaboration, imagine working on your automation flows with the same fluidity as Google Docs or Figma, where multiple collaborators can work on the same flow simultaneously, all within an intuitive and powerful interface. Our old canvas served us well for over five years, but it has become the main limiting factor for the new era. Like all node-based tools today, we are switching to React Flow, which will enable us to create much richer experiences. We're moving at a rapid pace to keep up with today's AI trends, and the new Designer will also support prompt-based development, similar to the Cursor experience mentioned earlier. All the updates we have been working on are geared toward this main generative flow experience. We’ve been releasing updates quickly, and the Designer has come a long way, we hope to publish it within a month. Follow us on our Discord where we share teasers.
🚀 New Admin:
We are revamping our administrative panel to provide a more consistent UI experience across all our updated products.
🚀 New Forms:
Our previous Forms development relied on a JSON object within your flows using the Form package. Now we are transitioning to a smart and efficient drag-and-drop interface. We are introducing dynamic forms that elevate user interactions to the next level. These forms can be created via drag and drop and include custom widget support similar to that available in the new Assistant. This allows you to transform complex tasks, such as automating ticket purchases, into a seamless, one-click experience. Additionally, you can use the Forms feature in various ways to collect data from your customers. Think of it as TypeForm, but dynamically created from your flow and shareable with anyone through your automation. We plan to release this feature after our new Designer release next month.
Here is an example of what was possible with the old format: https://forms.robomotion.io/#/XSqBat96Zs32b13MQMR65f
─────────────────────────────
AI vs. RPA: THE PERFECT COMBO
─────────────────────────────
While the AI landscape is evolving so rapidly that one year can feel like a decade, many of today's AI computer automation tools still face significant limitations. Models such as OpenAI Operator and Claude for Computer Use might impress in controlled demos, but they struggle with long-running tasks, operate slowly, and produce non-deterministic outputs. Computer Use achieves an accuracy of 38.1% in OS World (operating system navigation) and 58.1% in Web Arena (website navigation). Taking screenshots and determining the exact coordinates to click on your desktop in order to continue an automation is not reliable enough for a business environment, even for major AI providers. In contrast, RPA remains deterministic, reliable, and 100% consistent. Its strength lies in its precision: once an automation works, it works every single time.
By combining RPA’s reliability with AI’s adaptive intelligence, we harness the best of both worlds. This perfect combo empowers you to build automation flows that are both robust and smart enough to evolve with your needs.
─────────────────────────────
A HEARTFELT THANK YOU
─────────────────────────────
Thank you once again to our community of Sumo-lings. Your invaluable feedback has helped shape Robomotion into what it is today. We strongly believe that the future of automation lies in the integration of RPA and AI, and we're grateful to have your support as we continue to innovate.
Join our Discord at https://community.robomotion.io
Let’s build the future of automation together.
Thank you,
Faik Uygur
Co-Founder, Robomotion (edited)
