Frequently Asked Questions

In May 2025, Ymnig AI rebranded to Abundly AI.

The transition

Ymnig AI officially became Abundly AI in May 2025. This was a complete rebrand of the company while maintaining all the same services, team, and technology.

What this means

• Same company: All the expertise and people remain unchanged
• Same platform: Your AI agents and data continue to work seamlessly
• Same location: Still based in Stockholm, Sweden
• New identity: Now operating under the Abundly AI brand, more internationally recognizable and pronounceable but still aligned with our vision of creating abundance through AI technology

For existing customers

Nothing changes for your day-to-day experience. All accounts, agents, and services continue without interruption.

The rebrand was announced on May 28, 2025, and you can read more about it on our blog.

Abundly.ai is a startup based in Stockholm, Sweden, that multiplies human ability with AI agents.

Our mission

We help organizations achieve more through a combination of:

• AI Agent Platform: A comprehensive platform for deploying and managing AI agents
• Professional Services: Expert consulting and implementation support
• Training & Education: Programs to support organizations on their Gen AI journey

Our approach

We focus on practical AI implementation that delivers real business value. Our team combines deep technical expertise with business acumen to ensure successful AI adoption.

Location and focus

Based in Stockholm, Sweden, we serve organizations globally with a focus on making AI accessible and effective for businesses of all sizes.

Getting started with Abundly AI is straightforward. Here's how to begin your journey:

Create your account

1. Visit our signup page
2. Choose the plan that best fits your needs
3. Complete the registration process

Explore the platform

Once you have access, create an agent and start experimenting. Start with simple use cases before moving to more complex workflows.

Get help when needed

As you get started:

• Check our FAQ for common questions
• Reach out to support@abundly.ai if you need assistance
• Consider scheduling a demo or consultation if you have specific requirements

Pausing unused agents is one of the most effective ways to optimize your credit usage on Abundly.

How to pause an agent

1. Navigate to the agent you want to pause
2. Go to Settings within that specific agent
3. Find the Agent Status section
4. Toggle the switch from "Running" to "Off"

When to pause agents

• Automatically running agents that aren't being actively used
• Seasonal or project-specific agents that are temporarily not needed
• Testing or development agents when not in active development
• Backup agents that only need to run occasionally

What happens when paused

• Agent stops consuming credits immediately
• All configurations remain intact - nothing is lost
• Easy to restart - simply toggle the switch back to "Running"
• No data loss - all documents, chat history, and settings are preserved

Best practices

• Regular review: Periodically review your active agents and pause unused ones
• Monitor usage: Keep track of which agents consume the most credits
• Strategic pausing: Pause agents during off-hours or weekends if appropriate

Need help?

If you're unsure which agents to pause or need help optimizing your setup, contact us at support@abundly.ai.

Yes, our platform supports several users accessing the same agents – it's even something we recommend when re-engineering processes with the help of agents.

Access rights can be set per agent, starting from completely private agents to agents accessible to the entire team.

The recommended approach for fine-tuning AI agents is to:

1. Chat directly with the agent and update instructions with more specific details
2. Be very explicit about requirements (companies, time periods, etc.)
3. Try this method for 1-2 attempts
4. If you don't see improvement after 1-2 tries, contact Abundly support for additional assistance

This iterative approach of refining instructions through direct interaction is generally effective for improving agent performance.

Common issues and solutions

Date handling problems: If your agent includes wrong dates or information from outside specified time periods, this is often caused by a known issue where Claude (the underlying language model) can "hallucinate" dates when they are not explicitly provided.

Current solutions:

• Be very explicit with date ranges in your instructions
• Double-check that today's date is clearly specified in agent instructions
• Abundly is working on a platform-level fix to automatically provide dates to agents

When to escalate

If you've tried updating instructions 1-2 times and still experience issues:

1. Contact Abundly support with specific examples of the problems
2. Include details about what the agent should vs. shouldn't include
3. Provide examples of incorrect outputs
4. Support can help identify technical issues or provide advanced troubleshooting

Best practices for effective instructions

• Be very explicit about criteria (specific companies, date ranges, content types)
• Include examples of what should and shouldn't be included
• Specify the current date clearly in instructions
• Test changes iteratively rather than making many changes at once
• Document what works and what doesn't for future reference

To let your agent communicate on your slack server, you need to:

1. Create a Slack App and obtain a bot token and an App ID.
2. Enable the Slack capability in Abundly and paste in those two values.

Step 1: Create a slack app

Here is a simple quick-start guide. For more comprehensive info, see api.slack.com/docs.

The easiest way to create a slack app is to use a manifest file as a template. Copy the template below,

{
   "display_information":{
      "name":"YOUR_APP_NAME",
      "description":"YOUR_APP_DESCRIPTION",
      "background_color":"#000000"
   },
   "features":{
      "app_home":{
         "messages_tab_enabled":true,
         "messages_tab_read_only_enabled":false
      },
      "bot_user":{
         "display_name":"YOUR_BOT_DISPLAY_NAME",
         "always_online":true
      }
   },
   "oauth_config":{
      "scopes":{
         "bot":[
            "app_mentions:read",
            "channels:read",
            "chat:write",
            "groups:read",
            "im:read",
            "users:read",
            "channels:history",
            "groups:history",
            "im:history",
            "channels:join",
            "im:write",
            "mpim:history"
         ]
      }
   },
   "settings":{
      "event_subscriptions":{
         "request_url":"https://agent-service-339911427090.europe-north2.run.app/api/slack/webhook",
         "bot_events":[
            "app_mention",
            "message.im",
            "message.mpim"
         ]
      },
      "org_deploy_enabled":false,
      "socket_mode_enabled":false,
      "token_rotation_enabled":false
   }
}

1. Visit api.slack.com/apps
2. Click "Create New App"
3. Choose "From a manifest"
4. Paste the above manifest
5. Replace YOUR_APP_NAME, YOUR_APP_DESCRIPTION, and YOUR_BOT_DISPLAY_NAME. You can use your agent name for these.
6. OPTIONAL: adjust the scopes and bot_events if you like.

• The scopes define what the agent can do in slack. Here is the full list of scopes.
• The bot_events define which types of events will cause the agent to be triggered. Without bot_events the agent can only send and not receive messages on slack. Here the full list of events.
• Tip: With the configuration above, your agent will only be triggered if at-mentioned. If you want the agent to be triggered for all slack messages in the channels that it is a member of, then add message.channels and message.groups to bot_events.
• NOTE: The request_url link will probably change in the future. If you notice that your agent no longer receives slack messages, check back on this page.

5. Click "Next" and "Create". Your app is now installed!
6. Copy the App ID.
7. Go to "Oauth & Permission" and press the installation button under "OAuth Tokens". After doing so you should see "Bot User OAuth Token". Copy that value as well.

Step 2. Configure the slack capability in Abundly

Go to Settings -> Capabilities, enable slack, and paste in the bot token and app ID.

• token: this is the "Bot User OAuth Token" that you copied in the last step.
• appId: this is the "App ID" that you copied in the last step.

Done! You should now be able to interact with your agent on slack.

No, Abundly is not integrated with the Microsoft ecosystem and does not use Microsoft's selection of language models. We make independent choices about which AI models to use based on performance for autonomous agent tasks.

Overview

The GitHub integration enables AI agents to interact with GitHub repositories for pull request management and repository monitoring. This capability streamlines development workflows by automating common GitHub tasks.

What the GitHub Integration Can Do

Pull Request Management:

• Merge pull requests with different merge methods (merge, squash, rebase)
• List and filter pull requests by status
• Get detailed pull request information including file changes and diffs

Repository Access:

• Browse repository files and directories
• Read file contents from any branch or commit
• List repository branches
• Track recent commits and changes

Monitoring:

• Monitor repository activity
• Track changes since specific commits or dates
• Access all repositories you have permissions for

GitHub Projects Support

The GitHub integration does not currently support GitHub Projects. It focuses on repository-level operations like pull requests, commits, and file management.

Common Use Cases

• Automatically merge approved pull requests
• Generate release notes from commit history
• Monitor repository activity and send notifications
• Support code review workflows
• Track changes for release management

Getting Started

1. Enable the Capability: Go to Settings → Capabilities and enable "GitHub"

2. Create a GitHub Token:

   • Visit GitHub Settings → Personal access tokens
   • Click "Generate new token (classic)"
   • Select these scopes: repo, read:org, and workflow (if needed)
   • Copy the token immediately

3. Configure:

   • Paste your token in the "token" field
   • Optionally specify repositories in "watchedRepositories" (e.g., yourorg/project1)
   • Click "Save github Credential"

We use language models from providers such as Anthropic and OpenAI. These models do not store or remember any data from your sessions. According to their public documentation, user data is not retained or used for model training unless you explicitly opt in.

Please see our list of subprocessors.

Yes, as of June 2025, you need either an active subscription or sufficient credit balance to keep your agents running.

Subscription requirements

If you currently have a subscription:

• Your agents will continue running as long as you have credits
• You'll need to purchase additional credits when your balance runs low
• Your subscription gives you access to buy credits and keep agents active

If you don't have a subscription:

• You'll need to choose a subscription plan to continue using agents
• Without a subscription, agents will pause when credits are depleted
• You can select a plan that fits your usage needs and budget

Choosing the right plan

• Evaluate your usage: Consider how many agents you run and how frequently
• Review plan options: Check the available subscription tiers at app.abundly.ai/signup
• Start appropriately: You can always upgrade or downgrade as your needs change

What happens without a subscription

• Agents pause automatically when credits run out
• No data loss - all your work is preserved
• Easy reactivation - choose a plan and agents resume immediately

Need help choosing?

If you're unsure which subscription plan is right for you, or have questions about credit usage, contact our team at support@abundly.ai. We're happy to help you find the best fit for your needs.

There are several levers that affect how much it costs to run your AI agent. Here's what you can control:

1. Frequency of Use

How often your agent runs is a major factor. Triggering it every few minutes versus once an hour can lead to a substantial difference in monthly cost.

2. Size of the Context

Feeding the agent too much data — especially unnecessary context — quickly adds up. Trimming the context to only include what's essential is one of the most effective ways to cut costs without sacrificing quality.

3. Where You Store Data

This is an often-overlooked but powerful design choice. If you embed data directly into the agent's instructions, that data is included in every single run — even when it's not needed. By contrast, if you store supporting information in documents or tools that the agent can selectively retrieve, that data is only included when relevant. This can dramatically reduce the tokens used and save credit over time.

4. Complexity of Instructions

Recursive or multi-step instructions — where the agent performs several tasks or calls itself — can drive up usage significantly. Simplifying the agent's logic and reducing unnecessary loops can help manage costs effectively.

5. Language Model Selection

Currently, we use Claude 4 as our default model because we've found it provides the best balance of performance versus cost and is the easiest to get agents working reliably with. However, model choice can significantly impact costs.

Current Approach

We prioritize getting agents to work effectively first, then optimize costs later.

Future Options

We're exploring the possibility for users to choose different models based on their needs:

• Cheaper models for simpler tasks
• More expensive models for complex requirements
• Alternative models based on specific use cases

Cost Optimization Support

We're happy to help with cost optimization on a case-by-case basis, but haven't yet enabled universal model selection for all users.

Your Feedback is Welcome

This is an evolving area - let us know your thoughts and requirements around model selection and cost optimization.

Our agents run on the highest-performing language models available because that's what's required for them to function effectively. Cheaper models often fall short — they may struggle with long contexts and are more likely to ignore or override important instructions.

For simpler use cases, this can make models like Claude 3.5 Sonnet or similar appear more cost-efficient, but the trade-off in reliability, accuracy, and ability to follow complex instructions is significant.

Currently, your agent is powered by Claude 4.0 Sonnet. While there is a more expensive model — Claude 4.0 Opus — we've found that the performance difference does not justify the additional cost for most use cases. That's why we stick with Sonnet as our top-tier default.

Our Strategy

Start with the best model to ensure your agent performs reliably. Then, if the cost becomes too high for the frequency you'd like to run the agent, let us know which agent you're referring to and we'll be happy to downgrade it to a more cost-effective model.

Our AI agent platform is subscription-based, designed to scale with your organization's needs.

Current pricing

You can view the current pricing structure and plans on our signup page.

What's included

Our platform subscription typically includes:

• Access to the AI agents platform
• Basic support and documentation
• Regular platform updates and improvements

Custom solutions

For enterprise customers or organizations with specific requirements, we also offer custom pricing plans. Contact our team to discuss your specific needs.

Starting in June 2025, Abundly enforces credit limits to ensure sustainable platform usage.

When credits are depleted

• Agents automatically pause: Your agents will stop running when your credit balance reaches zero
• No interruption to data: All your agent configurations, documents, and chat history remain safe
• Easy restart: Agents resume immediately once you add more credits

What you need to do

If you have an active subscription:

• Purchase additional credits to keep your agents running
• Credits can be bought through your account dashboard

If you don't have a subscription:

• Choose a subscription plan to continue using your agents
• Visit your account settings to select the plan that fits your needs

Preventing unexpected stops

• Monitor your credit balance regularly in your dashboard
• Set up low-credit notifications if available
• Consider pausing unused agents to conserve credits

Getting help

If you need assistance with credit management or have questions about subscription plans, contact us at support@abundly.ai.

Yes, the Abundly team identified and fixed several recurring issues with scheduled tasks that were causing some scheduled agent activities to not run as expected.

What was fixed

• Multiple recurring issues that prevented scheduled tasks from running properly
• All scheduled tasks should now be running as expected
• The fixes were implemented and tested by the development team

When it was fixed

• The fixes were implemented and went live on the morning of June 26, 2025
• All scheduled tasks should be functioning normally as of that time

If you're still experiencing issues

• Contact the support team at support@abundly.ai if you notice any scheduled tasks still not working properly
• Include details about which agent and what scheduled tasks are affected

The 'Overloaded' error occurs when Anthropic/Claude's servers are experiencing high traffic and are temporarily overloaded.

What this means

This error happens when:

• Anthropic/Claude's servers are at capacity
• There's high demand on the AI model infrastructure
• The servers are temporarily unable to process new requests

How to resolve it

Wait and retry:

• This is a temporary issue that usually resolves within 5-10 minutes
• Simply try creating your agent again after waiting a few minutes
• The error typically clears up on its own as server load decreases

If the problem persists:

• Wait 10-15 minutes and try again
• If you're still experiencing issues after 15+ minutes, contact our support team at support@abundly.ai

Why this happens

This error is on Anthropic's side (the AI model provider), not with Abundly's platform. It's similar to when a popular website gets too much traffic - the servers need a moment to catch up with demand.

Future improvements

We're working on making this error message clearer and more user-friendly to better explain what's happening and what to do next.

For technical support and assistance, we offer multiple ways to get help.

Email support

For technical support, please email our team at support@abundly.ai.

What to include

When contacting support, please include:

• A clear description of the issue
• Steps you have already tried
• Your account information (if applicable)
• Screenshots or error messages (if relevant)

Response time

We typically respond to support requests within 24 hours during business days. For urgent issues, please mention "URGENT" in your subject line.

Currently, you cannot choose the language model in Abundly. We use Claude Sonnet 4.0 as it works best for our use cases. We continuously switch to whichever model performs best at any given time, ensuring optimal performance for autonomous agents.

Yes! As of June 2025, the Abundly team has implemented significant improvements that reduce credit consumption by 25-50% across all agent interactions.

What changed

• Reduced credit consumption for all agent activities (chat, autonomous tasks, scheduled operations)
• The exact reduction varies based on how interactions take place, but most users see 25-50% savings
• This improvement affects all types of agent usage regardless of context

When it took effect

• The changes have been live on the platform for several days as of June 26, 2025
• You do not need to take any action to benefit from these improvements
• All existing agents automatically benefit from the reduced credit consumption

User action required

None - This improvement is automatic and applies to all agents immediately. No configuration changes or updates are needed.

Chat conversations with our AI agents are stored in our own MongoDB database, which is hosted on a server in Stockholm, Sweden. These conversations are temporary and fully under user control. You can delete a conversation at any time, and once deleted, it is permanently removed from our servers.

Uploaded documents remain in the user's control and are not moved to our servers unless explicitly handled that way. For example, if you share a Google Doc with an agent, the document remains on Google's servers and is accessed only through your permission. The data stays wherever the original platform (e.g., Google) stores it.

Static files are served by a CDN, meaning is is dependent on user location. Our database, service and functions are run in Sweden.

Absolutely! You'll be very positively surprised. The concepts, strategies, and workflows you learn in our agent courses are valuable regardless of the specific AI platform you're using. While the implementation details may differ, the foundational knowledge about agent design, automation patterns, and AI integration principles applies broadly.