AI Workers With Their Own Computers, Accounts, and Budgets | Solid

AI workers with their own computers, accounts, and budgets.

Close your laptop. They will ping you when it's done.

Builders and operators at these companies use Solid.

-

-

-

-

-

-

-

-

-

-

-

-

-

-

- -

Hand over the job. Get the result.

You bring the goal and the ground rules. They work out the steps, check the result, and report back.

Update my dashboard when...

Tell them what you need.

Describe the job in plain language. They'll work out the steps.

Solid is working

  1. Connected to Gmail and HubSpot
  2. Researched on LinkedIn
  3. Built the dashboard
  4. Deployed the dashboard
  5. Verifying the data

Close your laptop. They keep working.

They connect to any tool the job needs, build what's missing, and check the result. No prebuilt connectors required.

Solid now

Your dashboard now updates automatically. Here's the link:

https://dashboard.solid.site

Finished work, delivered.

They message you with the result, ready to use or share.

They handle the setup and troubleshooting, too.

They work things out along the way, so you don't have to manage every step.

Self-sufficient

They get the tools and handle the setup.

They choose the tools and handle the setup, even for software you've never used. If a person can use it, they can too. No prebuilt connector required.

Self-healing

They get themselves unstuck.

If a tool fails or their setup breaks, they can investigate, make a repair, and check that the job runs again. When they need your help, they explain what is blocking progress.

Self-improving

The next job starts with what they learned.

They keep the fixes that worked and learn from your team's corrections. Those lessons change how they use tools and approach future jobs.

Self-scaling

They bring in help and manage it.

They create more Solid workers or bring in agents like Codex and Claude Code. They divide the work across any tools the job needs, coordinate the team, and return one checked result.

See what a finished job looks like.

Example workflows, from the first request to the result you can review.

They turn the customer's requirements into a tested demo and return the hosted link in your conversation.

Meeting notes become a demo you can send.

Turn these customer meeting notes into a sales demo. Use sample data, test the buyer workflow, and send me the hosted link. Ask before buying any services.

Sales demo worker

Example workflow

  1. Read the meeting notes
  2. Map the buyer's workflow
  3. Build the demo with sample data
  4. Host and test the app
  5. Return the link and test results

They research relevant accounts, prepare outreach, and carry approved conversations through follow-up and CRM updates.

Buying signals become outreach you can approve.

Research leads on LinkedIn using our ideal customer profile. Show me the evidence and draft outreach for approval. Follow up on approved messages, book qualified meetings, and update the CRM.

Lead generation worker

Example workflow

  1. Apply your targeting criteria
  2. Research buying signals
  3. Qualify accounts and contacts
  4. Draft outreach and wait for approval
  5. Follow up, book meetings, and update the CRM

They build evaluations that simulate customers using your AI product, judge the results, and send a detailed report after each release or change.

Know what works after every release.

Build and run evals for our AI product after every release or change. Simulate users completing key tasks, judge the results against the expected behavior, and send a report of what passed, what failed, and why.

AI evaluation worker

Example workflow

  1. Build user scenarios and success criteria
  2. Run after each release or change
  3. Simulate users completing real tasks
  4. Judge outcomes against expected behavior
  5. Report what passed, failed, and why

They investigate the issue, prepare a fix, and follow the work through deployment checks and customer updates.

An alert becomes a tested fix for review.

Investigate this production alert, identify affected customers, and open a tested fix for review. Wait for engineer approval before deployment, then verify recovery and prepare customer updates.

Bug resolution worker

Example workflow

  1. Investigate the production alert
  2. Reproduce the issue and assess its impact
  3. Write and test a fix
  4. Open a pull request for engineer approval
  5. Verify recovery after deployment

They trace the problem across your systems, complete an approved fix, and follow up with the customer.

A stalled ticket becomes a resolved issue.

Investigate this stalled support ticket using the full customer history. Complete the fix our policies allow, ask me about exceptions, and confirm the result with the customer.

Support worker

Example workflow

  1. Read the stalled ticket
  2. Gather the full customer history
  3. Find the cause across systems
  4. Apply the fix within your policies
  5. Confirm and record the resolution

Pricing

Your full monthly payment becomes one balance for AI usage, machines, and purchases your workers make. No extra platform fee.

$40/ month

For getting started with a focused task, a simple app, or a small workflow.

$160/ month

For regular work, active app building, and more room to test and iterate.

$640/ month

For heavier workloads, larger apps, and several projects running at once.

Start with $20 on us.

Use the trial to see what your work costs before choosing a plan.

Ship an AI worker in your product, or as the product.

Use the Solid API to deploy always-on agents inside your product. They can keep context, work across approved systems, build missing pieces, and carry long-running jobs under the controls you set.

Enterprise

Delegate more work without giving up control.

Set access, budgets, policies, and approvals across your workspace. Run on Solid Cloud, in your own VPC, or on-premises. Availability depends on your setup.

Before you hand over the first job.

What should I try first?

Start with one job whose result you can check, such as a sales demo, a researched lead list, or a weekly report. Describe what you want back. If you're unsure where to start, tell Solid what's taking up your time and ask what its workers could take on.

Do I need to set up the tools myself?

No. Start with the result you want and any useful files or instructions. Solid's workers work out which tools they need and can handle setup, connections, and code. You provide any required access and choose the approval rules.

Can they use my tools without a prebuilt connector?

Yes. Solid's workers can use any tool a person can use, with the access you approve. They can connect through an API, build a missing integration, or operate the website, desktop software, or phone app directly. You don't need to find your app on a connector list or know how to use it yourself. Describe the job, and they'll figure out the tools and setup it needs.

How do costs and spending limits work?

Solid has three monthly subscriptions: Starter at $40, Pro at $160, and Max at $640. The full amount is added to your account balance each month. AI usage, infrastructure, and approved purchases your workers make all draw from that same balance. We don't charge anything on top. You set the budgets and approval rules. Your worker is a meta-agent: it can see and manage its own workspace within the access you give it. Ask what’s running, why it’s needed, or how much a job cost, any time. It can break down the AI, machines, and purchases used for the job.

What can they do without my approval?

You choose what Solid's workers can access and which actions need your approval. For example, let them research and draft while requiring approval before sending a message, buying a service, or deploying a change. They work within the boundaries you set.

Do I need to stay online while they work?

No. Solid's workers can keep going after you close your laptop. They check the result and message you when it's ready, or ask when they need a decision. You can also check progress and review their work along the way.

What if they get stuck or make a mistake?

Solid's workers check their work and can revise it from your feedback. When a tool or setup fails, they can investigate, make a repair, and check that the job runs again. If they remain blocked, they explain what needs your help. You can also talk to a real person on the Solid team.

Will I have to explain everything again next time?

No. Solid's workers remember your team's instructions and feedback across jobs. When you correct how something is done, they keep that lesson and apply it the next time it's relevant.