RAiN AI Studio — the AI agent studio

RAiN AI Studio is the studio where RAiN experts design, run, and refine AI agents — the working engine behind the RAiN packages businesses buy. It lives at studio.rain.com.ng and uses your normal RAiN login: sign in once on RAiN and you're signed in everywhere.

Who it's for

RAiN AI Studio is an expert tool. It's available to verified RAiN experts (and the RAiN team); business accounts don't use the studio — they receive the finished solutions experts deliver with it.

Academy learners get in one other way: courses with a hands-on RAiN AI Studio lab open the studio with a practice workspace — the real tool, with the built-in Claude models, a tighter daily run limit, and complete separation from experts' client work.

How to open it

Sign in at www.rain.com.ng as an expert.
In your dashboard sidebar, under Grow, click RAiN AI Studio.
The studio opens in a new tab — no second login needed.

To come back, click the Dashboard button at the top-right of any Build page.

A tour of the studio

Area	What it's for
Studio	Your home: the prompt box that builds projects for you, and Your RAiN AI Projects below it
Templates	Ready-made agent setups from the RAiN package catalog — clone one and tailor it
LLM Connections	Which AI models your agents can use, and your own provider keys if you add any
Tools	The built-in abilities agents can be given (more connectors are on the way)
Dashboard (top-right)	Back to your RAiN account

Your first project, step by step

On the Studio home, describe what you want in the prompt box — for example: "A support agent that answers customer questions from our knowledge base, and escalates anything it can't answer." (Or click an example chip to start from one.)
Click Build it. RAiN designs the agents and tasks for you and opens them on a canvas.
Look around the workspace: your agents and tasks are connected on the Visual Editor canvas. Drag them, or click one to edit it.
Click Run, type a test input (say, a customer's question), and watch each agent's step appear live as it works.
When the run finishes, the full trace opens in the Execution tab — every step, the final output, and any errors.
Keep refining: use the Build with AI panel on the left to ask for changes in plain language ("add a quality-check step before the reply goes out").

Prefer starting from something proven? Open Templates and click Use template — it copies a RAiN package setup into your workspace, ready to customize.

The three tabs of a project

Visual Editor — the canvas: who the agents are, what the steps are, what follows what, and what starts it.
Execution — every run this project has done, with the full trace of each.
Settings — the project's name and description, its knowledge documents, who its messages come from, what it reports about every run, and the endpoint a client's system can call.

What a project is made of

Agents — the AI workers. Each has a role (who it is), a goal (what it's trying to achieve), an optional backstory (how it behaves), a model, and tools. An agent can perform a task written for it, or be handed work directly — in which case its role and goal are the instruction.
Steps — what happens, in order. A task when you want judgement, an action when you want certainty, a condition when it depends, an agents route when it should go to whichever specialist handles it.
Triggers — what starts a run. A clock, or the client's own system calling in. Without one, a project only runs when you press Run.
Knowledge — text and documents your agents can answer from: business facts, price lists, policies. Set it under Settings.

FAQs

Do I need to know how to code? No. You describe what you want in plain language; the studio designs and runs everything. Editing is point-and-click on the canvas.

I opened studio.rain.com.ng and see a sign-in screen. Sign in on www.rain.com.ng first, then return — one login covers the whole network. If you're signed in and still blocked, your account may not be an expert account yet.

Can my clients see my projects? No. Projects are private to your workspace — nobody else can open the canvas. What you can give a client is a way to run it: see Endpoint, where their own system calls your Sequence with a key you issue.

Can I delete a project? Yes — hover its card on the Studio home, click the trash icon, and confirm. Deleting is permanent and removes the project's run history too.

Is there a limit to how much I can run? Yes, a generous daily run limit (it resets at midnight UTC). See Running agents & reading traces.