Get Started with Noah
Try Noah with sample data — no setup required.
You’re one of the first people to use Noah. We’ve prepared a few sample datasets so you can explore the full workflow from process description to executable output in minutes.
See Noah in action
Watch this short video to see Noah in action, then try it yourself with the sample data below.
Three steps to your first process
Step 1 – Describe what you want to build
Tell Noah about your workflow in plain language. No forms, no templates. Just describe it the way you’d explain it to a colleague – Noah will structure it into a step-by-step process map as you go.
Step 2 – Upload the sample files
Download one of the sample datasets below and upload it when you start explaining your process. Noah will analyse the file structure, propose column mappings, and ask you the right questions before proceeding.
Step 3 – Review, validate, and deliver
Work through a short series of targeted questions to confirm the logic, mappings, and validation rules. Noah then runs a simulation on your data and packages the final output – ready to download with a full audit trail.
Download a sample dataset and start building
Each dataset comes with a brief process brief – a short description of what the workflow should do. Use it as your starting prompt when you open Noah.
Customer Master Reconciliation – 3 System Exports
What’s in the files: Three exports from different business systems – a CRM customer list, an accounting system pullout, and a support ticket export. Each system stores customer identity differently: some use a customer number, others rely on email or name and postal code.
The challenge: The files don’t share a consistent key. Names are stored in one field in some systems and split across first name and last name in others. Records exist in one system but not the others. Your job is to match them up, check for inconsistencies, and flag what needs a human decision.
Your starting prompt for Noah:
“Hi Noah! I want to create a new process to reconcile customer records from three sources: a CRM export, an accounting system pullout, and a support tickets export. Match records using customer number first, then email, then full name plus postal code as a fallback. For each customer, check whether they exist in all three sources, whether their contact data is consistent, and whether their status fields are logically aligned. Flag duplicates and ambiguous matches separately.”
Download 3 sample data files:
- CMR Reconciliation - 01_CRM_Kundenexport.xlsx
- CMR Reconciliation - 02_Accounting_System_Pullout.xlsx
- CMR Reconciliation - 03_Support_Tickets_Export.xlsx
A few things worth knowing before you start
- This is a beta. Noah works best when you describe your process in your own words – the more specific you are, the better the process map. You don’t need to get it right on the first message; you can refine as you go.
- Noah will ask you between 5 and 15 targeted questions before finalising any phase. These aren’t forms – they’re decision points where your input shapes the logic. Most questions have a recommended answer you can simply confirm when you agree with it.
- To start a fresh session run “Clean slate” command and then specify what Noah should do. Current options include:
- Process design – “I want to create a new process...”
- Process execution – “I want to execute my CRM reconciliation process...”
Help us make Noah better
You’re using an early version of Noah. If something feels off – a question that doesn’t make sense, a step that breaks, a result that looks wrong – we want to know.
Use the feedback button inside Noah, or drop us a note directly at feedback@celdon.ai. Every report you send directly shapes what we fix next.
.svg)
