ETUNIM AI Studio
One platform. Endless possibilities. Your agents. Your workflow. Your rules.
An environment in design for configuring, connecting and controlling multiple AI agents in automated workflows.
Last updated:
Description
ETUNIM AI Studio is a product being designed by ETUNIM. The aim is to bring specialised tools, task automation and collaboration between artificial intelligence agents together in a single environment. The idea is to let multiple agents share tasks and run coordinated workflows, recognising and correcting errors, with the user always in control.
This is not just another AI tool. We want it to be possible to choose the right agents for each task, define how they collaborate and follow what they do, from a single development environment.
The project is at the concept stage and is not yet available. Everything described on this page is what we intend to build.
Objectives
- Let multiple AI agents collaborate in a single workflow.
- Automate repetitive tasks, increasing efficiency and reducing time spent.
- Keep users in control of their agents, workflows and rules.
How we expect it to work
The planned experience, which may change as the project develops.
Choose your tools
Users choose the AI agents and tools suited to the task they want to carry out.
Connect your agents
They define how the agents connect, how they communicate and what information they share.
Customise your workflow
They set the instructions, task sequence, priorities, conditions and expected results.
Run and optimise
They run the workflow, follow the results and adjust the configuration whenever needed.
Workflow examples
Illustrative examplesThey show the idea behind the project. They are not demonstrations or real results, because the product does not exist yet.
Preparing a weekly report
- One agent gathers information from the chosen sources.
- Another summarises the main points.
- A third compares the summary with the defined rules and flags inconsistencies.
- Human approval
- A person reviews the report before it is sent.
- Intended result
- A report ready to send, with every step visible.
Reviewing a software change
- One agent analyses what changed in the code.
- Another runs the checks defined by the team.
- A third collects the errors found and proposes fixes.
- Human approval
- The team decides whether to apply the proposed fixes.
- Intended result
- An organised review, handed to the team for a decision.
Sorting requests received by email
- One agent classifies each request by subject and priority.
- Another looks for the information needed to reply.
- A third prepares a draft reply.
- Human approval
- No reply is sent without a person’s approval.
- Intended result
- Organised requests and drafts ready for review.
Control and monitoring
PlannedFollow every run
An automated workflow should not be a closed box. We want it to be possible to understand what happened and why.
- See the data passed between steps.
- Identify the step where an error occurred.
- Adjust the configuration based on what was observed.
Human approvals
Automation handles the routine, while important decisions stay with people.
- Define which actions need approval.
- Stop the workflow until a person decides.
- Review the work before it has effects outside the workflow.
On your phone too
The work should not depend on being at a computer.
- Follow runs in progress.
- See errors as they come up.
- Respond to approval requests remotely.
Planned features
PlannedThese features are planned and not yet available.
- Choose the AI agents and tools suited to each task.
- Connect agents: define how they communicate and share information.
- Customise the workflow: instructions, task sequence, priorities, conditions and expected results.
- Run and optimise, following results and adjusting the configuration.
- Execution monitoring, showing data passed between steps and identifying errors.
- Human approvals before important actions.
- Follow the work from a phone, not only from a computer.
Project roadmap
- In progressProductM-020
Defining the AI Studio concept
Defining the four-step user experience: choose tools, connect agents, customise the workflow, run and optimise.
- Priority
- High
- Target date
- Date to be confirmed
- Last updated
- PlannedProductM-021
Real workflow demonstrations
Publish examples of tasks, agent connections and real results, once available.
- Priority
- Medium
- Target date
- Date to be confirmed
- Last updated
Frequently asked questions
Is ETUNIM AI Studio available yet?
No. The project is at the concept stage and has no launch date. You can ask to be notified of updates using the email option on this page.
Which AI agents and models will it support?
This has not been defined yet. Technologies will be listed on this page once confirmed.
Will I need to know how to code?
The final experience is still being defined. The idea includes the option of coding workflow logic, but whether this will be required has not been decided.
How much will it cost?
No prices or plans have been defined yet.
Other projects
SoftwareConcept
Business Management
Web applications and management software for small and medium-sized businesses (SMEs), at the concept stage, to simplify processes and automate tasks.
Game developmentConcept
Independent Games
Independent development of video games and interactive experiences, as well as tools that support game creation, at the concept stage.